Skip to content
View in the app

A better way to browse. Learn more.

hosang I.T.

A full-screen app on your home screen with push notifications, badges and more.

To install this app on iOS and iPadOS
  1. Tap the Share icon in Safari
  2. Scroll the menu and tap Add to Home Screen.
  3. Tap Add in the top-right corner.
To install this app on Android
  1. Tap the 3-dot menu (⋮) in the top-right corner of the browser.
  2. Tap Add to Home screen or Install app.
  3. Confirm by tapping Install.

Tech

Tech Articles from a wide variety of topics and categories
In einem falschen Security-Mindset gefangen?
Foto: Paul Craft – shutterstock.com
Dass Jobs im Bereich Cybersecurity ein hohes Burnout-Potenzial aufweisen, ist längst kein Geheimnis mehr: Das Umfeld von Sicherheitsprofis ist vor allem geprägt von dem (gefühlten) Druck, täglich steigenden Anforderungen gerecht werden zu müssen. Dafür sind diverse Gründe ursächlich – in erster Linie aber die Art und Weise, wie über Security gedacht wird. Die gute Nachricht: Wenn Sie ein schädliches Mindset identifizieren, können Sie es verändern und sowohl sich als auch Ihre Teams besser für den Erfolg positionieren.
Cybersicherheit ist ein hochtechnisches Gebiet und in gewisser Hinsicht eine harte Wissenschaft. Auf der anderen Seite ist sie aber auch stark von Elementen der Psychologie und Moral geprägt. Wie effektiv die IT-Sicherheit letztlich ausfällt, hängt auch vom Mindset und den Überzeugungen der Fachkräfte und Entscheider auf diesem Gebiet ab.

Sollten Sie eines der folgenden sechs Mindsets an den Tag legen, ist Arbeit angesagt, damit ein gesünderes Security-Umfeld gedeihen kann.
1. “Security ist ein Ziel”
Ein besonders heimtückisches Security-Mindset ist die Überzeugung, dass es sich um eine Reise mit Start- und Zielpunkt handelt. Zu dieser Überzeugung kommt man (hoffentlich) nicht bewusst – Profis ist klar, dass es sich um eine kontinuierliche Aufgabe handelt. Unterbewusst kann es aber durchaus dazu kommen, dass es zu vorübergehender Untätigkeit kommt, wenn bestimmte Tasks gerade erledigt wurden.
Das führt allerdings nur dazu, dass alle im Team mehr unnötigen Stress haben. Denn wer ein Ende in Aussicht stellt, erzeugt ein subtiles Gefühl der Enttäuschung oder gar des Scheiterns, sobald offenbar wird, dass es doch immer noch etwas mehr zu tun gibt. Zur Ruhe werden Sie (und Ihr Team) erst kommen, wenn sie akzeptieren, dass Security ein fortlaufender Prozess ist.
2. “IT-Sicherheit ist nur was für Profis”
Die Auffassung, dass Security ausschließlich in den Händen der entsprechenden Spezialisten liegt, führt zu zweierlei unglücklichen Konsequenzen:
Alle anderen Mitarbeiter werden – zumindest gefühlt – aus der Verantwortung entlassen.
Sicherheitsprofis werden auf subtile Weise in eine Einzelkämpferrolle gedrängt.
Softwareentwickler sollten Security in jeder Phase des Lebenszyklus im Hinterkopf behalten, statt sich erst zur Auslieferung damit zu befassen. Das gilt jedoch auch für alle anderen Mitarbeiter im Unternehmen: Nur wenn Awareness herrscht, kann die Gefahr von Cyberangriffen minimiert werden.
Natürlich kommt den Sicherheitsexperten diesbezüglich eine führende, beziehungsweise leitende Rolle zu. Letztendlich sollte sich aber jeder Mitarbeiter dazu befähigt fühlen, zur allgemeinen Unternehmenssicherheit beitragen zu können. Eine gemeinschaftliche Aufgabe stärkt davon abgesehen auch das Wir-Gefühl.
3. “Security wird immer nur diffiziler”
Kaum etwas ist entmutigender als eine klassische Sisyphos-Aufgabe. Dieser Eindruck kann allerdings leicht entstehen, wenn es um Security geht: Cyberkriminelle werden immer raffinierter und nutzen immer bessere Tools, während die digitale Infrastruktur, die geschützt werden muss, sich immer umfangreicher, komplexer und vernetzter gestaltet.
In der Realität ist der Kampf zwischen White und Black Hats ein ständiges Geben und Nehmen. Das Phänomen Ransomware ist ein gutes Beispiel: Eine Zeit lang schienen sich Verschlüsselungstrojaner zu einer Plage zu entwickeln – inzwischen hat sich die Sicherheitsbranche entsprechend weiterentwickelt und messbar zurückgeschlagen.
Indem Sie die zyklische Natur der IT-Sicherheit akzeptieren, befähigen Sie sich dazu, eine Haltung einzunehmen, die die richtige Balance zwischen Entspannung und Wachsamkeit findet. Mentales Gleichgewicht ist der Schlüssel zu langfristigem (Security-)Erfolg.
Sie wollen weitere interessante Beiträge rund um das Thema IT-Sicherheit lesen? Unser kostenloser Newsletter liefert Ihnen alles, was Sicherheitsentscheider und -experten wissen sollten, direkt in Ihre Inbox.
Jetzt CSO-Newsletter sichern
4. “Sicherheit ist ein Produkt”
Die IT Security wird nicht selten als Standalone-Funktion oder Zusatzprodukt betrachtet, die über die zugrundeliegende Infrastruktur “gestülpt” wird oder als konkrete “Sache”, die finalisiert und ausgeliefert werden muss. Das ähnelt ein bisschen der einstigen Perspektive auf Qualität im Allgemeinen als eine eigenständige, separate Komponente der Dinge. Um es mit Aristoteles zu sagen: “Qualität ist keine Handlung, sondern eine Gewohnheit”.
Security wiederum ist wie Qualität kein fertiges Produkt, sondern (wie bereits angemerkt) eine fortlaufende Disziplin. Sicherheit als eine Praxis zu betrachten, die ständig verfeinert werden muss, setzt die dafür nötige Energie frei. Sie sollten es als Segen betrachten, in einem Bereich zu arbeiten, der kontinuierlich Raum für Wachstum und die Möglichkeit bietet, Ihre Skills vollumfänglich zur Geltung zu bringen. Haben Sie dieses Mindset verinnerlicht, gilt es, das mit dem gesamten Unternehmen zu teilen.
Security sollte in keinem Fall wie ein Produkt ausgeliefert werden, denn sie ist keine Begleiterscheinung oder ein Hilfsmittel. Vielmehr sollte sie der Treiber für Kultur und bewusstes Handeln sein. Kurzum: IT-Sicherheit sollte Teil des täglichen Doings sein – auf individueller und organisatorischer Ebene.
5. “Die Kriminellen treiben die Security”
Security-Profis, die kontinuierlich damit beschäftigt sind, Brände zu löschen, können zur Überzeugung kommen, dass die Cyberkriminellen das Spiel beherrschen. Diese reaktive Perspektive auf die IT-Sicherheit sorgt für Frustration und ein Gefühl der Machtlosigkeit.
In der Realität haben die Unternehmen das Ruder in der Hand: Sie sind es schließlich, deren Assets für Kriminelle verlockende Ziele darstellen. Die Angreifer sind in den meisten Fällen nicht zu unterschätzen – es ist jedoch das Business, dass die Sicherheit treibt.
6. “100 Prozent reicht gerade”
Gute Sicherheit braucht messbare Faktoren. Metriken wie die “Mean Time to Detect” (MTTD) ermöglichen es, die Situation zu monitoren und die Effektivität von Programmen zu messen. Problematisch wird es in diesem Bereich, wenn Sie der Vorstellung erliegen, dass sich sämtliche Indikatoren stets in eine positive Richtung – oder noch schlimmer im “perfekten” Bereich – bewegen müssen. Diese unrealistische Erwartung ist ein Einfallstor für verzerrte Messwerte.
Stattdessen sollten Sie Metriken eher als Wegweiser sehen, die Sie ans Ziel bringen können. Der Schlüssel liegt jedoch darin, die nötigen Schritte zu unternehmen und Maßnahmen einzuziehen, um die Dinge in die richtige Richtung lenken. Das macht es essenziell, sich ehrlich mit Messungen auseinanderzusetzen. (fm)



View the full article
A recently disclosed high-severity security flaw in Apache ActiveMQ Classic has come under active exploitation in the wild, per the U.S. Cybersecurity and Infrastructure Security Agency (CISA). To that end, the agency has added the vulnerability, tracked as CVE-2026-34197 (CVSS score: 8.8), to its Known Exploited Vulnerabilities (KEV) catalog, requiring Federal CivilianView the full article
Customers are more loyal to Apple than ever, according to a smartphone loyalty survey conducted by phone trade-in site SellCell. 96.4% of customers surveyed said they planned to stick with an iPhone for their next upgrade, and 3.6% said they would choose a different brand. That's up from 91.9% in SellCell's 2021 survey and 90.5% in 2019.


Android users were less loyal to their brand, and are almost 4x more likely to switch than iPhone users. 86.4% of people surveyed said they would stick with an Android device, while 13.6% said they planned to switch.

Of the 3.6% of iPhone users who said they would move to another platform, 69.7% said they would choose a Samsung smartphone, and 20.2% said they would choose a Google smartphone. While most Android users said they would switch to a Samsung or Google device, 26.8% said they would choose an iPhone over an Android smartphone.

Most iPhone users said they would stick with an iPhone because they prefer Apple (60.8%), while 17.4% said they were invested in the Apple ecosystem. About half of iPhone users contemplating switching said they would do so because the iPhone is too expensive or other brands offer better value, but 22.5% said other brands have better technology.

iPhone users were more likely to be loyal over time, and 83.8% said they had used an iPhone for more than five years. By comparison, just 33.8% of Android users said they had stuck with a brand for over five years.

SellCell's survey was limited to 5,000 U.S. respondents. The site says there was a roughly equal representation between iPhone and Android users, with two separate surveys that included the same question structure. More information from the survey is available from SellCell's website.Tags: Android, SellCell
This article, "iPhone Loyalty Hits 96.4% as Android Users Four Times More Likely to Switch" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Apple hardware engineering chief John Ternus and marketing chief Greg Joswiak recently did an interview with Tom's Guide, where they shared new insights into the MacBook Neo, AI, and spatial computing.


Ternus and Joswiak made it clear that the ‌MacBook Neo‌ isn't your average low-cost device. Apple doesn't typically put a lot of focus on its more affordable devices, but marketing for the Neo has been expansive, and that's because Apple sees it as a "reinvention" of the entry-level laptop. From Ternus:

Ternus said the ‌MacBook Neo‌ required "building something completely new from the ground up" to provide customers with quality at a low price. "We never want to ship junk," he said. "We want to ship great products that have that Apple experience."

Joswiak said the ‌MacBook Neo‌'s quality was important to Apple, and the Neo's build sets it apart from competitors.

Along with discussing the Neo, Ternus and Joswiak talked about the differences between the iPad and the Mac. Ternus said that Apple isn't going to merge the products, and similarities are because Apple focuses on what would make a device better and not on how one product might impact another.

On AI, which is an area where Apple has been struggling, Joswiak said it's not a sprint.

Joswiak dodged a question about a potential touchscreen MacBook Pro, which Apple is rumored to be working on for launch as soon as this year. He also declined to comment on smart glasses, but said we're in the "early innings of spatial computing," while Ternus said that combining the digital and physical world is an "inevitability." The two were tight-lipped about any upcoming Apple products, but Joswiak said Apple is "working on some pretty cool stuff."

The full interview, which goes into more detail on the ‌MacBook Neo‌, AI, and includes a Steve Jobs anecdote, is well worth watching.Tags: Greg Joswiak, John Ternus
This article, "Apple Execs Say Spatial Computing Is 'Inevitable' and AI Is a 'Marathon, Not a Sprint'" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Admins who use Cisco Webex Services configured to use trust anchors within the SSO integration with Control Hub must install a new identity provider certificate to close a critical vulnerability, or risk losing access control.
Cisco said in an advisory this week that admins must upload a new identity provider (IdP) SAML certificate to Webex Control Hub, the web-based management portal where IT administrators can control all Cisco Webex services, including certificate management, meetings, messaging and calling. Failure to close this hole will allow an unauthenticated, remote attacker to impersonate any user within the service.
The vulnerability, CVE-2026-20184, carries a CVSS score of 9.8.
Because Webex is a cloud service, Cisco can, and has, patched its side of the application. But admins using single-sign on (SSO) still need to install the new certificate. There are no workarounds.
A Webex support article on managing SSO integration says that information about certificates is found in the Webex Control Hub Alerts center, where customers can view which ones are installed, and their status. The Control Hub also contains an SSO wizard to aid in updating certificates. The article contains step-by-step details on the process.
Asked for comment, and for more details about the vulnerability, a Cisco spokesperson didn’t go beyond the advisory. “Cisco published a security advisory disclosing a vulnerability in the integration of single sign-on with Control Hub in Cisco Webex Services,” the spokesperson said. “At the time of publication (April 15) Cisco had addressed the vulnerability, and was not aware of any malicious use of this vulnerability. Affected customers must update their SAML certificate to ensure uninterrupted services.”
Gartner analyst Peter Firstbrook noted in an email that, since Cisco has applied the patch to the cloud service, this is more of a configuration change. But that doesn’t minimize the possible damage. “While we are not aware of exploits using this vulnerability, users can lose SSO access to Webex without this change,” he said. 
“This does illustrate a bigger trend that identity and access management is the corporate perimeter,” he added, “and the majority of attacks include an identity and access management component. CISOs must increase their focus on IAM hygiene, particularly as agentic computing is accelerating.” 
Identity and access management is, of course, the keystone of cybersecurity. As Crowdstrike observed in its 2026 Global Threat Report, abuse of valid accounts accounted for 35% of cloud incidents it investigated last year, “reinforcing that identity has become central to intrusion.” Single sign-on allows a user to authenticate to multiple applications through one set of credentials. It’s efficient, and, of more importance to a CSO, strengthens security.
Additional critical fixes
The Webex flaw is one of three critical vulnerabilities Cisco identified and issued patches for this week. In addition, multiple vulnerabilities have to be patched in Cisco Identity Services Engine (ISE) and Cisco ISE Passive Identity Connector (ISE-PIC).
These holes (CVE-2026-20147 and CVE-2026-20148, which carry CVSS scores of 9.9), could allow an authenticated, remote attacker to perform remote code execution or conduct path traversal attacks on an affected device. To exploit these vulnerabilities, the attacker must have valid administrative credentials, and send a crafted HTTP request to an affected device. There are no workarounds.
Separately, two more vulnerabilities were found in ISE that could lead to remote code execution on the underlying operating system of an affected device. To exploit these vulnerabilities (CVE-2026-20180 and CVE-2026-20186), the attacker would only need Read Only Admin credentials.

View the full article
Admins who use Cisco Webex Services configured to use trust anchors within the SSO integration with Control Hub must install a new identity provider certificate to close a critical vulnerability, or risk losing access control.
Cisco said in an advisory this week that admins must upload a new identity provider (IdP) SAML certificate to Webex Control Hub, the web-based management portal where IT administrators can control all Cisco Webex services, including certificate management, meetings, messaging and calling. Failure to close this hole will allow an unauthenticated, remote attacker to impersonate any user within the service.
The vulnerability, CVE-2026-20184, carries a CVSS score of 9.8.
Because Webex is a cloud service, Cisco can, and has, patched its side of the application. But admins using single-sign on (SSO) still need to install the new certificate. There are no workarounds.
A Webex support article on managing SSO integration says that information about certificates is found in the Webex Control Hub Alerts center, where customers can view which ones are installed, and their status. The Control Hub also contains an SSO wizard to aid in updating certificates. The article contains step-by-step details on the process.
Asked for comment, and for more details about the vulnerability, a Cisco spokesperson didn’t go beyond the advisory. “Cisco published a security advisory disclosing a vulnerability in the integration of single sign-on with Control Hub in Cisco Webex Services,” the spokesperson said. “At the time of publication (April 15) Cisco had addressed the vulnerability, and was not aware of any malicious use of this vulnerability. Affected customers must update their SAML certificate to ensure uninterrupted services.”
Gartner analyst Peter Firstbrook noted in an email that, since Cisco has applied the patch to the cloud service, this is more of a configuration change. But that doesn’t minimize the possible damage. “While we are not aware of exploits using this vulnerability, users can lose SSO access to Webex without this change,” he said. 
“This does illustrate a bigger trend that identity and access management is the corporate perimeter,” he added, “and the majority of attacks include an identity and access management component. CISOs must increase their focus on IAM hygiene, particularly as agentic computing is accelerating.” 
Identity and access management is, of course, the keystone of cybersecurity. As Crowdstrike observed in its 2026 Global Threat Report, abuse of valid accounts accounted for 35% of cloud incidents it investigated last year, “reinforcing that identity has become central to intrusion.” Single sign-on allows a user to authenticate to multiple applications through one set of credentials. It’s efficient, and, of more importance to a CSO, strengthens security.
Additional critical fixes
The Webex flaw is one of three critical vulnerabilities Cisco identified and issued patches for this week. In addition, multiple vulnerabilities have to be patched in Cisco Identity Services Engine (ISE) and Cisco ISE Passive Identity Connector (ISE-PIC).
These holes (CVE-2026-20147 and CVE-2026-20148, which carry CVSS scores of 9.9), could allow an authenticated, remote attacker to perform remote code execution or conduct path traversal attacks on an affected device. To exploit these vulnerabilities, the attacker must have valid administrative credentials, and send a crafted HTTP request to an affected device. There are no workarounds.
Separately, two more vulnerabilities were found in ISE that could lead to remote code execution on the underlying operating system of an affected device. To exploit these vulnerabilities (CVE-2026-20180 and CVE-2026-20186), the attacker would only need Read Only Admin credentials.

View the full article
AI agent building tools enable users to configure Model Context Protocol (MCP) servers may be exposing systems to remote code execution due to an architectural decision in Anthropic’s reference implementation.
At issue are unsafe defaults in how MCP configuration works over the STDIO interface, with broad implications for the agent ecosystem, according to a new report.
“The blast radius is massive,” researchers from application security firm OX Security wrote in their report on the design issue. “This exploit allowed us to directly execute commands on six official services of real companies with real paying customers, and to take over thousands of public servers spanning over 200 popular open-source GitHub projects with hundreds of millions of downloads.”
According to Anthropic and other MCP adapter developers, the STDIO command execution behavior is by design and the responsibility of sanitizing MCP configurations falls with developers of client applications. While this might be true, in practice OX Security found that few developers have attempted to filter commands in MCP configs and even those who did failed to catch all potential bypasses.
The root of the issue
MCP provides a standardized method for applications to expose data sources and tools to LLMs, improving their context and effectiveness in completing automated workflows. Originally developed by Anthropic, MCP has become a widely adopted technology in the agentic AI space.
Anthropic provides reference MCP implementations in the form of SDKs for a variety of programming languages, including TypeScript, Python, Java, Kotlin, C#, Go, PHP, Ruby, Rust, and Swift. Furthermore, other frameworks and functionality providers — such as FastMCP, LangChain’s mcp-adapters, Microsoft’s agent-framework, mcp-agent, browser-use, Amazon’s run-model-context-protocol-servers-with-aws-lambda, and NVIDIA’s NeMo-Agent-Toolkit — have Anthropic’s modelcontextprotocol reference implementation as a dependency.
MCP supports two transport interfaces between servers and clients: Streamable HTTP with Server-Sent Events (SSE), which is typically used for remote MCP servers and web services, and Standard Input/Output (STDIO), for MCP servers and applications that run locally on the same machine.
With STDIO, client applications can start MCP servers on demand as a subprocess and pass parameters to them. These parameters can include custom commands that get executed on the system with the permissions of the parent process. While in theory these commands are meant to tell the SDK’s StdioServerParameters function how to start the MCP server, they can technically be anything if no filtering is in place.
The OX Security researchers consider this a design flaw that should be mitigated, but Anthropic disagrees, as do the creators of other frameworks that enable MCP functionality, such as LangChain and FastMCP. The argument is that the responsibility for making sure malicious user input doesn’t reach the SDK’s command execution function resides with the developers of the client applications that integrate these MCP frameworks.
“The pattern of allowing user-supplied strings to flow directly into a shell execution environment is an anti-pattern that should be deprecated,” the OX Security researchers said. Anthropic’s SDKs should implement a command allowlist by default that blocks sh, bash, powershell, curl, rm, and other high-risk binaries, they added.
The core issue is that there’s currently no check in place to verify that a STDIO command is intended to initialize an MCP server rather than perform a malicious task. Furthermore, the researchers observed that even if the sent command fails to start the server, the SDK returns an error after the command has already been executed.
All modern IDEs such as VS Code, Cursor, and Windsurf, as well as agentic coding CLIs like Claude Code, OpenAI Codex, and Gemini CLI, have built-in support for local MCP servers over STDIO. But so do countless other agentic AI frameworks and open-source tools and few of them implement STDIO command allow lists.
RCE in real-world applications
The OX Security researchers have spent the past few months testing MCP support in numerous tools, including live production services. They found and reported more than 30 RCE issues stemming from this STDIO design decision to multiple projects and 10 have received CVE IDs so far.
Depending on how a tool implements MCP support and how it accepts user input, there are multiple attack vectors that exploit the lack of STDIO command filtering.
For example, some services and tools have not disabled STDIO internally even though their user interfaces only allow configuring MCP servers with Streamable HTTP. This was the case for Letta AI and DocsGPT, two platforms that enable companies to create AI agents via both cloud services and local deployments.
“An attacker crafting a network request for an MCP server configuration, and changing the transport type in the configured JSON to contain an STDIO type instead of SSE or HTTP, also adding an arbitrary command to the request’s payload, can achieve remote command execution,” the researchers said.
Another attack vector is prompt injection leading to malicious MCP configurations. While all IDEs are technically vulnerable to this — websites may contain hidden instructions for LLM agents to modify local files — most IDEs prompt users before making modifications to MCP configuration files. The exception was Windsurf, which directly modified the MCP config by default, resulting in a zero-interaction command injection attack.
Many other tools don’t apply filtering to MCP STDIO parameters, meaning any user with access to configure an MCP server gains code execution on the underlying server, including production servers in the case of SaaS deployments. Tools found vulnerable to this include LangFlow, GPT Researcher, LiteLLM, Agent Zero, LangBot, Fay Digital Human Framework, Bisheng, Jaaz, Langchain-Chatchat, and several others the researchers are not yet able to disclose.
Some developers were aware of the issue and did attempt to harden their implementations with command whitelisting. However, the hardening was insufficient, and the OX Security researchers found simple bypasses.
For example, Upsonic, an open-source framework for building AI agents, implements an allowlist that includes npx, which supports -c (—call), a flag that allows custom commands and shell scripts to be passed for npx to execute. The same bypass was observed in Flowise, another UI-based AI agent building framework that also restricts MCP configuration commands but allows npx.
Anthropic (modelcontextprotocol), LangChain (langchain-mcp-adapters), FastMCP, the browser-use project, AWS (run-model-context-protocol-servers-with-aws-lambda), NVIDIA (NeMo-Agent-Toolkit), OpenHands, PromptFoo, Firebase Studio, Gemini CLI, Claude Code, GitHub Copilot, and Cursor technically include the MCP STDIO code that allows for arbitrary command execution.
Their maitainers consider the command execution intended behavior or have declined to fix it because it is mitigated by other controls such as modifications requiring user interaction or the command execution happening inside sandboxed environments like Docker containers.
See also:
What CISOs need to know about new tools for securing MCP servers Top 10 MCP vulnerabilities: The hidden risks of AI integrations 6 ways attackers abuse AI services to hack your business View the full article
Overwhelmed by an escalating volume of security flaws, the National Institute of Standards and Technology (NIST) has announced significant changes to how it handles cybersecurity vulnerabilities and exposures (CVEs).
Rather than commit to providing enrichment for all entries in its National Vulnerability Database (NVD), the agency will focus on just the most critical CVEs, which will “allow us to stabilize the program while we develop the automated systems and workflow enhancements required for long-term sustainability.”
Starting immediately, NIST will focus on CVEs appearing in CISA’s Known Exploited Vulnerabilities (KEV) catalog. “Our goal is to enrich these within one business day of receipt,” the agency said.
Other high-priority CVEs will also include those for software used in the federal government and for other critical software.
All the other CVEs will still be added to the NVD, but will be categorized as “not scheduled,” meaning that NIST will no longer prioritize their enrichment.
Broken by backlog
According to NIST, a backlog of CVEs started to accumulate in early 2024, and the agency has been unable to clear it due to increasing submissions.
Submissions grew by 263% between 2020 and 2025, according to the agency, with nearly one-third more vulnerabilities reported in Q1 2026 than the same time last year.
The agency, which enriched nearly 42,000 CVEs in 2025, 45% more than any previous year, now faces a total backlog of more than 30,000 CVEs, said Harold Booth, a technical and program lead at NIST, at this week’s VulnCon cybersecurity conference.
SOURCE: https://www.cve.org/about/Metrics
CSO
As a result, NIST will now forego enrichment for all but the most critical of vulnerabilities.
Backlogged CVEs received prior to March 1 will also be labeled “not scheduled.” None of those are critical vulnerabilities, NIST said, because those have always been handled first.
“They’ve just come out and publicly stated, ‘We are never going to get through this backlog,’“ Dustin Childs, head of threat awareness at Trend Micro’s Zero Day Initiative, told CSO.
In addition, NIST will no longer calculate severity scores for CVEs submitted with scores provided by the reporting organization.
Security leaders reliant on NIST enrichment will need to take stock of their technology inventories to see whether they fall under NIST’s priority list, Childs said. That’s not easy.
“Discovery is one of the most difficult problems we’re dealing with,” he noted, adding that it’s also not clear what software actually falls into the priority category. “Software used by the federal government is a very vague statement.”
Mounting CVE counts — with AI flaw discovery on the rise
Childs is not surprised that CVEs numbers have been going up, citing AI as part of the reason why.
“We’re already seeing more garbage CVEs — and more real CVEs — related to AIs,” he says.
Dealing with these CVEs is going to be a massive problem for companies. “People still don’t patch,” he says. “And we’re going to quadruple the number of patches they’re going to have to deploy. How do we build our defenses across the entire enterprise? I don’t know if we’ll get there before the bad guys do.”
According to the Forum of Incident Response and Security Teams (FIRST), 59,427 CVEs are expected to be submitted this year, up from a little over 48,000 in 2025. That makes 2026 the first year that CVEs will pass the 50,000 milestone.
“The sheer velocity of vulnerability discovery and exploitation is unlike anything we’ve seen before,” FIRST CEO Chris Gibson told CSO.
FIRST has also modeled “realistic scenarios” in which the total number of CVEs cracks 100,000 for 2026 — but that was in February, before Anthropic announced Mythos, its vulnerability-finding AI model many foresee as a structural shift for the cybersecurity industry.
“And if it’s not Mythos, or whatever else is coming out now, something is going to come out next week,” said Empirical Security founder Jay Jacobs, who also leads the Exploit Prediction Scoring System special interest group at FIRST.
Still, Jacobs is optimistic that turning to technology will help NIST deal with rising CVE volumes.
“Harold Booth has a lot of experience and skill working with AI over the last few years,” Jacobs told CSO. “So I’m expecting him to bring some expertise and I hope we do see some AI news there.”
Both large language models and AI agents are on the agency’s to-do list, as is old-fashioned robotic process automation (RPA), Booth said in his presentation at VulnCon, which Jacobs chairs. NIST also plans to delegate some of the work to CVE Numbering Authorities (CNAs), which includes security vendors and researchers.
View the full article
Overwhelmed by an escalating volume of security flaws, the National Institute of Standards and Technology (NIST) has announced significant changes to how it handles cybersecurity vulnerabilities and exposures (CVEs).
Rather than commit to providing enrichment for all entries in its National Vulnerability Database (NVD), the agency will focus on just the most critical CVEs, which will “allow us to stabilize the program while we develop the automated systems and workflow enhancements required for long-term sustainability.”
Starting immediately, NIST will focus on CVEs appearing in CISA’s Known Exploited Vulnerabilities (KEV) catalog. “Our goal is to enrich these within one business day of receipt,” the agency said.
Other high-priority CVEs will also include those for software used in the federal government and for other critical software.
All the other CVEs will still be added to the NVD, but will be categorized as “not scheduled,” meaning that NIST will no longer prioritize their enrichment.
Broken by backlog
According to NIST, a backlog of CVEs started to accumulate in early 2024, and the agency has been unable to clear it due to increasing submissions.
Submissions grew by 263% between 2020 and 2025, according to the agency, with nearly one-third more vulnerabilities reported in Q1 2026 than the same time last year.
The agency, which enriched nearly 42,000 CVEs in 2025, 45% more than any previous year, now faces a total backlog of more than 30,000 CVEs, said Harold Booth, a technical and program lead at NIST, at this week’s VulnCon cybersecurity conference.
SOURCE: https://www.cve.org/about/Metrics
CSO
As a result, NIST will now forego enrichment for all but the most critical of vulnerabilities.
Backlogged CVEs received prior to March 1 will also be labeled “not scheduled.” None of those are critical vulnerabilities, NIST said, because those have always been handled first.
“They’ve just come out and publicly stated, ‘We are never going to get through this backlog,’“ Dustin Childs, head of threat awareness at Trend Micro’s Zero Day Initiative, told CSO.
In addition, NIST will no longer calculate severity scores for CVEs submitted with scores provided by the reporting organization.
Security leaders reliant on NIST enrichment will need to take stock of their technology inventories to see whether they fall under NIST’s priority list, Childs said. That’s not easy.
“Discovery is one of the most difficult problems we’re dealing with,” he noted, adding that it’s also not clear what software actually falls into the priority category. “Software used by the federal government is a very vague statement.”
Mounting CVE counts — with AI flaw discovery on the rise
Childs is not surprised that CVEs numbers have been going up, citing AI as part of the reason why.
“We’re already seeing more garbage CVEs — and more real CVEs — related to AIs,” he says.
Dealing with these CVEs is going to be a massive problem for companies. “People still don’t patch,” he says. “And we’re going to quadruple the number of patches they’re going to have to deploy. How do we build our defenses across the entire enterprise? I don’t know if we’ll get there before the bad guys do.”
According to the Forum of Incident Response and Security Teams (FIRST), 59,427 CVEs are expected to be submitted this year, up from a little over 48,000 in 2025. That makes 2026 the first year that CVEs will pass the 50,000 milestone.
“The sheer velocity of vulnerability discovery and exploitation is unlike anything we’ve seen before,” FIRST CEO Chris Gibson told CSO.
FIRST has also modeled “realistic scenarios” in which the total number of CVEs cracks 100,000 for 2026 — but that was in February, before Anthropic announced Mythos, its vulnerability-finding AI model many foresee as a structural shift for the cybersecurity industry.
“And if it’s not Mythos, or whatever else is coming out now, something is going to come out next week,” said Empirical Security founder Jay Jacobs, who also leads the Exploit Prediction Scoring System special interest group at FIRST.
Still, Jacobs is optimistic that turning to technology will help NIST deal with rising CVE volumes.
“Harold Booth has a lot of experience and skill working with AI over the last few years,” Jacobs told CSO. “So I’m expecting him to bring some expertise and I hope we do see some AI news there.”
Both large language models and AI agents are on the agency’s to-do list, as is old-fashioned robotic process automation (RPA), Booth said in his presentation at VulnCon, which Jacobs chairs. NIST also plans to delegate some of the work to CVE Numbering Authorities (CNAs), which includes security vendors and researchers.
“Among other things, we are pursuing efforts to determine how large language models and other machine learning tools can be leveraged to speed up analysis and enrichment tasks that are currently manual and labor-intensive,” Booth added in follow-up with CSO.
This story has been updated to include added comment from Harold Booth on NIST’s AI plans.
View the full article
iPhone accessory maker Casely reissued a recall for its faulty Power Pod wireless power bank (via The Verge) after one of the affected units resulted in the death of a 75-year-old woman and another exploded on a plane.


Casely first issued the recall in April 2025 through the Consumer Product Safety Commission. At that time, the company said that the power banks could overheat and ignite, posing a fire and burn hazard to consumers. A total of 51 people had reported incidents where the battery overheated, expanded, or caught fire, and there were six minor burn injuries.

Since then, there have been an additional 28 reports, including two serious incidents. In August 2024, a woman in New Jersey was charging her phone with a Casely power bank on her lap, and it caught fire and exploded. She had second- and third-degree burns, and later died from complications from her injuries.

In February 2026, a 47-year-old woman was charging her cell phone with the power bank on an airplane when it caught fire and exploded, resulting in first-degree burns. Airlines have introduced more restrictive limits on power banks due to incidents like this.

Casely sold 429,200 power banks, which were branded as the Casely "Power Pod" with MagSafe compatibility. The 5,000mAh wireless power banks were available in multiple colors and patterns, and were priced at between $30 and $70. Affected units have an E33A model number and were sold from Amazon.com, the Casely website, and other websites between March 2022 and September 2024.

Anyone with a Casely Power Pod should stop using it immediately and contact Casely for a free replacement or a $60 store credit. Affected units should not be discarded, and customers should contact their local household hazardous waste collection center for disposal assistance. Casely is contacting all known purchasers directly.Tag: MagSafe
This article, "Casely MagSafe-Compatible Power Banks Recalled Again After Fire-Related Death and In-Flight Explosion" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Apple's MacBook Neo has been a huge hit, and it's still in high demand over a month after it launched. The ‌MacBook Neo‌ is just $599, and with PC makers raising prices because of global RAM shortages, the Neo's low price tag and Apple allure are even more appealing.


‌MacBook Neo‌ orders placed today on the online Apple Store won't reach customers until May, which means that it's sold out for the month of April, as 9to5Mac points out. All colors and both the 256GB and 512GB SSD configurations will be delivered between May 1 and May 8 at the earliest.

Some Apple retail locations have in-store availability today in select colors, but other stores won't have stock until May 11. Third-party retailers don't appear to have immediate stock, with Best Buy and Target listing delivery dates at least a week out.

Demand has exceeded expectations, and Apple is ramping up production. Apple is now planning to ship 10 million units in 2026, up from the original five to six million estimate. After the ‌MacBook Neo‌ launched, Apple CEO Tim Cook said that Apple saw its "best launch week ever for first-time Mac customers."

Apple may need to refresh the ‌MacBook Neo‌ sooner than expected because Apple does not have an unlimited supply of the binned A18 Pro chips that the machine uses. Apple could run out of the A18 Pro chip before it is able to satisfy ‌MacBook Neo‌ demand. Apple may need to restart A18 Pro chip production, which has ended, or start using an A19 Pro chip instead.

We'll likely hear more about the ‌MacBook Neo‌'s success during Apple's April 30 earnings call for the second fiscal quarter of 2026.Related Roundup: MacBook NeoBuyer's Guide: MacBook Neo (Buy Now)Related Forum: MacBook Neo
This article, "Apple's $599 MacBook Neo Sold Out Through April Amid Surging Demand" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Perplexity today launched Personal Computer, an expansion of Perplexity Computer that integrates with local files and apps on a Mac. Personal Computer was announced in March and was available on a waitlist basis, but it is officially rolling out today for Max subscribers.


Perplexity Computer came out earlier this year, and it's an all-in-one "digital worker" able to create and execute entire workflows. With today's upgrade, it can run directly on a Mac with access to the file system and native apps. Pressing both Command keys on a Mac will activate Personal Computer, and it responds to text or voice commands. Personal Computer can work across any Mac app, and it can see active apps and display quick actions automatically.

Perplexity says Personal Computer can run on any Mac with macOS 14 Sonoma or later, but the company recommends a Mac mini. With a ‌Mac mini‌, Personal Computer can run 24/7 for work that requires a persistent machine or secure local access to files and native apps. Tasks can be initiated and managed from an iPhone on the go.

Personal Computer can do things like complete each task on a to-do list, sort a messy downloads folder, compare local files against information on the web, and more. It can create teams of agents across over 20 frontier models to complete tasks. Personal Computer's actions are visible, so users can step in when needed. Files are created in a secure sandbox, the actions that Personal Computer takes are auditable and reversible, and there is a kill switch.

Personal Computer for Mac is rolling out to Perplexity Max subscribers starting today, with Perplexity prioritizing waitlist members. Perplexity Max is priced at $200 per month, and the new feature is not available to $20/month Pro plan subscribers.Related Roundup: Mac miniTag: PerplexityBuyer's Guide: Mac Mini (Caution)Related Forum: Mac mini
This article, "Perplexity Launches Personal Computer for Mac, Turning a Mac mini Into an Always-On AI Agent" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Cybersecurity researchers have warned of an active malicious campaign that's targeting the workforce in the Czech Republic with a previously undocumented botnet dubbed PowMix since at least December 2025. "PowMix employs randomized command-and-control (C2) beaconing intervals, rather than persistent connection to the C2 server, to evade the network signature detections," Cisco TalosView the full article
Last week, we launched Docker Sandboxes with a bold goal: to deliver the strongest agent isolation in the market.
This post unpacks that claim, how microVMs enable it, and some of the architectural choices we made in this approach.
The Problem With Every Other Approach
Every sandboxing model asks you to give something up. We looked at the top four approaches.
Full VMs offer strong isolation, but general-purpose VMs weren’t designed for ephemeral, session-heavy agent workflows. Some VMs built for specific workloads can spin up more effectively on modern hardware, but the general-purpose VM experience (slow cold starts, heavy resource overhead) pushes developers toward skipping isolation entirely.
Containers are fast and are the way modern applications are built. But for an autonomous agent that needs to build and run its own Docker containers, which coding agents routinely do, you hit Docker-in-Docker, which requires elevated privileges that undermine the isolation you set up in the first place. Agents need a real Docker environment to do development work, and containers alone don’t give you that cleanly.
WASM / V8 isolates are fast to spin up, but the isolation model is fundamentally different. You’re running isolates, not operating systems. Even providers of isolate-based sandboxes have acknowledged that hardening V8 is difficult, and that security bugs in the V8 engine surface more frequently than in mature hypervisors. Beyond the security model, there’s a practical gap: your agent can’t install system packages or run arbitrary shell commands. For a coding agent that needs a real development environment, WASM isn’t one.
Not using any sandboxing is fast, obviously. It’s also a liability. One rm -rf, one leaked .env, one rogue network call, and the blast radius is your entire machine.
Why MicroVMs
Docker Sandboxes run each agent session inside a dedicated microVM with a private Docker daemon isolated by the VM boundary, and no path back to the host.
That one sentence contains three architectural decisions worth unpacking.
Dedicated microVM. Each sandbox gets its own kernel. It’s hardware-boundary isolation, the same kind you get from a full VM. A compromised or runaway agent can’t reach the host, other sandboxes, or anything outside its environment. If it tries to escape, it hits a wall.
Private, VM-isolated Docker daemon. This is the key differentiator for coding agents. AI is going to result in more container workloads, not fewer. Containers are how applications are developed, and agents need a Docker environment to do that development. Docker Sandboxes give each agent its own Docker daemon running inside a microVM, fully isolated by the VM boundary. Your agent gets full docker build, docker run, and docker compose support with no socket mounting, no host-level privileges, none of the security compromises other approaches require. This means we treat agents as we would a human developer, giving them a true developer environment so they can actually complete tasks across the SDLC.
No path back to the host. File access, network policies, and secrets are defined before the agent runs, not enforced by the agent itself. This is an important distinction. An LLM deciding its own security boundaries is not a security model. The bounding box has to come from infrastructure, not from a system prompt.
Why We Built a New VMM
Choosing microVMs was the easy part. Running them where developers actually work was the hard part.
We looked hard at existing options, but none of them were designed for what we needed. Firecracker, the most well-known microVM runtime, was designed for cloud infrastructure, specifically Linux/KVM environments like AWS Lambda. It has no native support for macOS or Windows, full stop. That’s fine for server-side workloads, but coding agents don’t run in the cloud. They run on developer laptops, across macOS, Windows, and Linux. 
We could have shimmed an existing VMM into working across platforms, creating translation layers on macOS and workarounds on Windows, but bolting cross-platform support onto a Linux-first VMM means fighting abstractions that were never designed for it. That’s how you end up with fragile, layered workarounds that break the “it just works” promise and create the friction that makes developers skip sandboxing altogether.
So we built a new VMM, purpose-built for where coding agents actually run.
It runs natively on all three platforms using each OS’s native hypervisor: Apple’s Hypervisor.framework, Windows Hypervisor Platform, and Linux KVM. A single codebase for three platforms and zero translation layers.
This matters because it means agents get kernel-level isolation optimized for each specific OS. Cold starts are fast because there’s no abstraction tax. A developer on a MacBook gets the same isolation guarantees and startup performance as a developer on a Linux workstation or a Windows machine.
Building a VMM from scratch is not a small undertaking. But the alternative, asking developers to accept slower starts, degraded compatibility, or platform-specific caveats, is exactly the kind of asterisk that makes people run agents on the host instead. Our approach removes that asterisk at the hypervisor level.
Fast Cold Starts
We rebuilt the virtualization layer from scratch, optimizing for fast spin up and fast tear downs. Cold starts are fast. This matters for one reason: if the sandbox is slow, developers skip it. Every friction point between “start agent” and “agent is running” is a reason to run on the host instead. With near-instant starts, there is no performance reason to run outside it.
What This Means In Practice
Here’s the concrete version of what this architecture gives you:
Full development environment. Agents can clone repos, install dependencies, run test suites, build Docker images, spin up multi-container services, and open pull requests, all inside the sandbox. Nothing is stubbed out or simulated. Agents are treated as developers and given what they need to complete tasks end to end. 
Scoped access, not all-or-nothing. You define the boundary: exactly which files and directories the agent can see, which network endpoints it can reach, and which secrets it receives. Credentials are injected at runtime and outside the MicroVM boundary, never baked into the environment.
Disposable by design. If an agent goes off track, delete the sandbox and start fresh in seconds. There is no state to clean up and nothing to roll back on your host.
Works with every major agent. Claude Code, Codex, OpenCode, GitHub Copilot, Gemini CLI, Kiro, Docker Agent, and next-generation autonomous systems like OpenClaw and NanoClaw. Same isolation, same speed, one sandbox model across all of them.
For Teams
Individual developers can install and run Docker Sandboxes today, standalone, no Docker Desktop license required. 
For teams that want centralized filesystem and network policies that can be enforced across an organization and scale sandboxed execution, get in touch to learn about enterprise deployment.
The Tradeoff That Isn’t
The pitch for sandboxing has always come with an asterisk: yes, it’s safer, but you’ll pay for it in speed, compatibility, or workflow friction.
MicroVMs eliminate that asterisk. You get VM-grade isolation with cold starts fast enough that there’s no reason to skip it, and full Docker support inside the sandbox. There is no tradeoff.
Your agents should be running autonomously. They just shouldn’t be running without any guardrails.
Use Sandboxes in Seconds
Install Sandboxes with a single command.
macOS
brew install docker/tap/sbx   
Windows
winget install Docker.sbx  
Read the docs to learn more.

View the full article
OpenAI is making several updates to its Codex AI coding agent. Codex is now able to operate desktop Mac apps with its own cursor, seeing what's on the screen, clicking, and typing to complete tasks.


Codex can run multiple agents on the Mac in parallel, without interfering with the user's own work. OpenAI says developers will find it useful for testing apps, iterating on frontend changes, and more. Codex can now remember preferences, recurring workflows, tech stacks, and other information about each user's personal workflow. With automation improvements, Codex is able to resume work after a pause using existing conversation threads, and it can schedule future work for itself and work on a task across days or weeks. Codex also proposes work using context from projects, memory, and connected plugins.

There is an in-app browser for Codex that allows users to comment directly on pages to provide more precise instructions to the agent. In the future, Codex will get full use of the browser for opening websites, working through user flows, taking screenshots, and inspecting outputs.

Codex has been updated to use gpt-image-1.5 for generating images in the app, which OpenAI says is helpful for creating visuals for product concepts and mockups. Codex now includes support for multiple terminal tabs, addressing GitHub review comments, and opening files directly in the sidebar with rich previews for documents like PDFs and spreadsheets.

Along with these changes, Codex has over 90 new plugins that can combine skills, app integrations, and MCP servers to improve Codex's context gathering and actions.

The updates to Codex are rolling out today to Codex desktop users signed in with ChatGPT. The personalization features are not yet available to Enterprise, Education, EU, and UK users, but will be rolling out soon. Computer use is also not yet available in the EU or the UK.Tag: OpenAI
This article, "OpenAI Codex Update Adds Computer Use, Image Generation, and Memory on Mac" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
In a video uploaded to its YouTube channel in South Korea today, Apple showed off a handful of iPhone 17 Pro devices decorated with tiny stickers.


The stickers are placed on the iPhone 17 Pro's so-called "plateau," the protruding aluminum area housing the rear cameras, an LED flash, a microphone, and the LiDAR Scanner. The video has the hashtags #PhoneDecor and #iPhoneCustomization.

"Stick it here," says Apple.


The ad is accompanied by a pair of YouTube Shorts.Related Roundup: iPhone 17 ProTags: Apple Ads, South KoreaBuyer's Guide: iPhone 17 Pro (Neutral)Related Forum: iPhone
This article, "Apple Covers iPhone 17 Pro in Stickers in New Ad" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Amazon recently introduced a few new record low prices on the M5 MacBook Air, offering $150 off multiple models of the notebook. We've begun noticing that deals aren't as plentiful as they were over the weekend, so if you've been holding off now will be the time to get the M5 MacBook Air at these best-ever prices.

Note: MacRumors is an affiliate partner with Amazon. When you click a link and make a purchase, we may receive a small payment, which helps us keep the site running.

Amazon has the 512GB 13-inch M5 MacBook Air for $949.00, down from $1,099.00, and the 24GB/1TB model for $1,349.00, down from $1,499.00. Both of these represent record low prices for each configuration.

$150 OFF13-inch M5 MacBook Air (512GB) for $949.00
$150 OFF13-inch M5 MacBook Air (16GB/1TB) for $1,149.00
$150 OFF13-inch M5 MacBook Air (24GB/1TB) for $1,349.00

In terms of the 15-inch models, you'll find up to $150 off the M5 MacBook Air, with multiple color options on sale for each configuration. Prices start at $1,149.00 for the 512GB model, down from $1,299.00, and also include both 1TB models on sale.

$150 OFF15-inch M5 MacBook Air (512GB) for $1,149.00
$150 OFF15-inch M5 MacBook Air (16GB/1TB) for $1,349.00
$150 OFF15-inch M5 MacBook Air (24GB/1TB) for $1,549.00

If you're on the hunt for more discounts, be sure to visit our Apple Deals roundup where we recap the best Apple-related bargains of the past week.



Deals Newsletter

Interested in hearing more about the best deals you can find in 2026? Sign up for our Deals Newsletter and we'll keep you updated so you don't miss the biggest deals of the season!




Related Roundup: Apple Deals
This article, "Get Apple's M5 MacBook Air at $150 Off Before Amazon Runs Out" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Apple is offering 10% off AirPods, Beats, or accessories to customers who recycle an eligible iPhone, iPad, Apple Watch, or Mac at a participating Apple Store through May 16.


The discount applies when the recycling and purchase are completed in the same transaction. The promotion runs from today, April 16, through to May 16. Products brought in for recycling undergo screenings, with eligible devices sent to Apple's advanced recycling systems, including Daisy and Cora, for further processing. See Apple's terms and conditions for more information.

The promotion was announced alongside Apple's annual Environmental Progress Report, which revealed a record 30% of material across all products shipped in 2025 came from recycled content.Related Roundup: AirPods 4Buyer's Guide: AirPods (Neutral)Related Forum: AirPods
This article, "Apple Offers 10% Off AirPods and More for Earth Day Device Recycling" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Apple today announced that a record 30% of material across all products it shipped in 2025 came from recycled content, alongside a series of other environmental milestones published in its annual Environmental Progress Report.


The achievement marks new highs across several specific components. All batteries designed by Apple now use 100% recycled cobalt, all magnets use 100% recycled rare earth elements, and all Apple-designed printed circuit boards use 100% recycled gold plating and tin soldering. Apple also completed the transition to fully fiber-based packaging, fulfilling a pledge to remove all plastic from packaging by 2025, a goal the company says it reached across every package manufactured today.

Apple's greenhouse gas emissions in 2025 remain down over 60% compared to 2015 levels, holding constant from 2024 despite significant business growth. The company is working toward its Apple 2030 goal of carbon neutrality across its entire footprint by the end of the decade.

MacBook Neo leads the lineup on recycled material. It contains 60% recycled content overall, which is the most of any Apple device to date, and features a new aluminum forming process that uses half the raw material compared to traditional machining. Apple and its suppliers also developed an anodization process that achieves a 70% water-reuse rate, turning a traditionally water-intensive step into a near-closed-loop system. Apple said it plans to expand this process to additional production lines in coming years.

Apple launched Cora, a new electronics-recycling line at its Advanced Recovery Center in California, designed to achieve material recovery rates significantly higher than industry baselines using precision shredding and advanced sensor technology. The company also developed A.R.I.S., a machine learning-powered detection system that helps recyclers classify and sort electronic scrap, running on the Mac mini, which Apple is piloting with partner recyclers.

Apple's direct suppliers procured more than 20 gigawatts of renewable energy in 2025 through the Supplier Clean Energy Program, generating more than 38 million megawatt-hours of electricity, which is enough to power more than 3.4 million U.S. households for a year. Apple itself procured an additional 1.8 gigawatts to power its offices, retail stores, and data centers entirely on renewable electricity.

Apple and its suppliers saved 17 billion gallons of fresh water in 2025, and the company replenished more than half of the water it withdrew to support its global facilities. All eight Apple-owned data centers have now been certified to the Alliance for Water Stewardship standard. Apple has set a goal to replenish all water withdrawn by its facilities worldwide by 2030.

Apple Fifth Avenue in New York City became the company's first retail store to achieve TRUE Zero Waste Certification, which requires facilities to divert more than 90 percent of their waste from landfills. Across its supply chain, Apple and its suppliers redirected more than 600,000 metric tons of waste from landfills in 2025, with 400 supplier facilities participating in the company's Zero Waste Program.Tag: Apple Environment
This article, "Apple Hits Record 30% Recycled Content Across All Products in 2025" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Microsoft’s Windows Recall feature remains vulnerable to complete data extraction despite a major security overhaul, according to a cybersecurity researcher who says malware running in a user’s context can quietly siphon off everything Recall has captured, without administrator privileges, kernel exploits, or breaking encryption.
Alexander Hagenah, executive director at Zürich-based financial infrastructure operator SIX Group, made the claim in a LinkedIn post, where he also published a proof-of-concept tool called TotalRecall Reloaded to demonstrate the issue.
Hagenah first exposed Recall’s security flaws in 2024, forcing Microsoft to pull the feature from preview and rebuild it. Microsoft relaunched Recall in April 2025, saying the new architecture would restrict “attempts by latent malware trying to ‘ride along’ with a user authentication to steal data.” Hagenah said it does not.
“When you use Recall normally, TotalRecall Reloaded silently holds the door open behind you and then extracts what Recall has ever captured. That is precisely the scenario Microsoft’s architecture is supposed to restrict,” he wrote in the post.
Hagenah wrote in the post that he disclosed the research to Microsoft’s Security Response Center on March 6, submitting full source code and reproduction steps. Microsoft reviewed the case for a month and closed it on April 3, telling him the behavior “does not represent a bypass of a security boundary or unauthorized access to data.”
“Microsoft says this is by design,” Hagenah wrote. “That worries me.”
Hagenah’s research does not challenge Microsoft’s encryption, which he said is sound. The gap, he told CSO, is in how decrypted data is handled once it leaves the enclave.
“Plaintext screenshots and extracted text end up in an unprotected process for display,” he told CSO. “As long as decrypted content crosses into a process that same-user code can access, someone will find a way in.”
What a fix would require
A fix is technically feasible, Hagenah said.
“The short-term fix is fairly straightforward. Microsoft could add stronger code integrity and process protections to AIXHost.exe, the process that renders the Recall timeline. Right now, it has none, which makes the injection path possible. That would block the specific technique I demonstrated and materially raise the bar,” he said.
The longer-term problem runs deeper, he said. “Microsoft should rethink how decrypted data is handled after it leaves the enclave. The cryptography and enclave design are genuinely well done, and I want to be clear about that. The problem is that plaintext screenshots and extracted text end up in an unprotected process for display. As long as decrypted content crosses into a process that same-user code can access, someone will find a way in,” he said.
“A durable fix would mean either rendering inside a protected process or adopting a compositing model where raw data never leaves the trust boundary. That is a bigger effort, but it is the only way to close this class of issue properly,” he said.
Exploitation risk
The barrier to weaponizing this technique is lower than Microsoft’s security messaging would suggest, Hagenah said.
“They only need code running in the user’s context and a way to reuse the authorized Recall session,” he said. “That is a much lower bar than many people would assume from Microsoft’s security messaging.”
While Recall’s limitation to Copilot+ PCs and its opt-in status reduce the scale of exposure, targeted abuse is a realistic near-term risk, he said. “For targeted abuse, surveillance, or high-value user collection, this is absolutely realistic,” he said.
Hagenah said he published the source code deliberately so defenders, EDR vendors, and security teams could build detections before threat actors operationalize the technique independently. “In my view, that gives the defensive side a valuable head start,” he said.
Independent security researcher Kevin Beaumont reached a similar conclusion after separately testing the current Recall implementation. “Yep, you can just read the database as a user process,” Beaumont wrote on Mastodon on March 11. “The database also contains all manner of fields that aren’t publicly disclosed for tracking the user’s activity. No AV or EDR alerts triggered,” he wrote.
Microsoft did not immediately respond to a request for comment.
View the full article
In 2024, compromised service accounts and forgotten API keys were behind 68% of cloud breaches. Not phishing. Not weak passwords. Unmanaged non-human identities that nobody was watching. For every employee in your org, there are 40 to 50 automated credentials: service accounts, API tokens, AI agent connections, andOAuth grants. When projects end or employees leave, mostView the full article
Efforts to cut through the buzz surrounding Anthropic’s Mythos are emerging. As OpenAI moves to counter the hype around it with its own cybersecurity model, VulnCheck is reporting that the model’s publicly attributable output amounts to just one confirmed CVE.
While Project Glasswing, the controlled access program for Mythos, promises a powerful offensive capability, gated behind vetted organizations, VulnCheck’s recent findings reveal what those capabilities actually represent in practice.
“Anthropic’s Project Glasswing has generated significant attention—but very little concrete data,” said Patrick Garrity, researcher at VulnCheck, in a blog post. “While Anthropic researchers are actively contributing to vulnerability discovery and appear to be promising, the publicly attributable impact of Glasswing itself remains limited so far.”
Anthropic did not immediately respond to CSO’s request for comments.
Only one CVE is attributable to Glasswing
VulnCheck’s analysis of Project Glasswing drills into the numbers behind the claims by looking into public CVE attribution. “I started by re-reading the Glasswing report and the advisories published at red.anthropic.com,” Garrity said. “Neither source provides a comprehensive CVE list of vulnerabilities discovered by Anthropic. So I decided to search the full CVE record database, and searched every CVE record containing the term “anthropic” and reviewed each one.”
Garrity identified 75 CVE records that mention Anthropic. But only 40 of those were actually credited to Anthropic researchers, with the rest tied to affected products or unrelated references. Of those 40, 10 originated from external collaboration programs, such as Calif.io’s MADBugs initiatives.
The 40 CVEs attributed to Anthropic researchers span multiple products, including 28 affecting Firefox, nine tied to wolfSSL, and one each impacting NGINX Plus, FreeBSD, and OpenSSL.
When narrowed down further, the number that mattered the most showed up. Only one CVE is explicitly attributed to Project Glasswing itself, CVE-2026-4747. This is a FreeBSD NFS remote code execution (RCE) flaw described as autonomously identified and exploited.
Garrity did not include the three vulnerabilities without CVE numbers mentioned on the Glasswing page. These include a 27-year-old OpenBSD flaw, a 16-year-old FFmpeg bug, and Linux kernel privilege escalation chains, all under embargo pending patches.
Why is Glasswing still a big deal
VulnCheck’s findings reframe Glasswing’s capabilities. The limited number of directly attributable CVEs is just one way of measuring its impact. Industry observers are interpreting Mythos much differently.
Melissa Bischoping, a SANS Technology Institute board member and senior Director of security and product research at Tanium, thinks Mythos potential lies elsewhere. According to a breakdown of the Claude Mythos Preview System Card, which Bischoping and her colleagues at Tanium reviewed, the model achieved an unseen exploit success rate. “Jumping from near-zero success to ~72% on the same class of targets suggests exploit development is no longer a high-skill, high-effort bottleneck,“ she said, adding that it’s only a matter of time before every other model catches up.
While Mythos is being regulated under Glasswing, it has already shown the world what is possible. “The gap between frontier models and open-weight models has compressed from more than a year to a matter of weeks, which means this level of capability is poised to spread rapidly, likely without the same safety guardrails,” Bischoping noted.
Bischoping is also concerned about whether organizations can act on what Mythos finds before Mythos is out in the wild. “Agentic patch workflows are possible and can match pace with adversarial AI in a lot of cases, but org politics and change control don’t run at the speed of AI today.”
The full picture about the model’s true capability won’t be known before July 2026, when Anthropic will make a full public accounting of what Glasswing found and fixed, Garrity said.
View the full article
A British paraplegic adventurer was being filmed for an Apple Vision Pro immersive video series during a fatal aircraft crash in the Jordanian desert in July 2024, Bloomberg's Mark Gurman reports.


Claire Lomas became internationally recognized in 2012 when she became the first person to complete the London Marathon using a robotic exoskeleton suit, five years after being paralyzed from the waist down in a horse-riding accident. Apple was apparently working with London-based Atlantic Studios to film Lomas for its Apple Immersive Video series called Adventure. A camera system was mounted on the plane itself, and Lomas was actively being recorded when the crash occurred. Lomas died from her injuries within weeks of the crash at the age of 44.

The planned episode was set to showcase multiple Jordanian landmarks, including the Wadi Rum valley and the ancient city of Petra. The original release date for was sometime in 2025. Apple spent millions of dollars per episode on the series, with Atlantic producing and Apple distributing.

People involved with the production say there were broader safety concerns during the making of the Adventure series, including crews working longer hours than staff felt was safe, filming in harsh climates, and operating equipment in conditions the crew had limited training for. Staffers reportedly raised these concerns with their superiors at Apple, which in response sent a health and safety representative to work periodically alongside production staff. There is no record of other major injuries tied to the series.

Apple and Atlantic continued working together after the crash; a Colorado episode was filmed in August 2024. Apple has released five Adventure episodes to date, featuring athletes highlining 3,000 feet in the air, swimming under Arctic ice, parkouring across Paris, cliff diving in Spain, and racing cars in Colorado. No new episodes have been published since last year.

The Adventure series sits within Apple's broader Apple Immersive Video offering for the Vision Pro, which Apple describes as a "180-degree, 3D 8K recording format captured with Spatial Audio." The format is one of the headset's primary selling points and is used in in-store demonstrations of the $3,499 device.Related Roundup: Apple Vision ProBuyer's Guide: Vision Pro (Buy Now)Related Forum: Apple Vision Pro
This article, "Apple Was Filming Fatal Aircraft Crash for Vision Pro" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Cisco has announced patches to address four critical security flaws impacting Identity Services and Webex Services that could result in arbitrary code execution and allow an attacker to impersonate any user within the service. The details of the vulnerabilities are below - CVE-2026-20184 (CVSS score: 9.8) - An improper certificate validation in the integration of single sign-on (SSO)View the full article
Introduction
Modern IT infrastructure has reached a level of complexity where manual intervention is no longer enough to maintain uptime. As environments scale, the volume of logs, metrics, and traces becomes overwhelming for human operators. This is where the Certified AIOps Professional program becomes essential for your career growth. This guide is designed for engineers and managers who want to move beyond traditional monitoring into the world of intelligent automation. By following the roadmap provided by AIOps School, professionals can learn how to apply machine learning and data science to IT operations. This guide will help you understand the certification landscape, evaluate the curriculum, and decide which path aligns with your long-term career goals in the global market.
What is the Certified AIOps Professional?
The Certified AIOps Professional is a practitioner-focused program designed to bridge the gap between traditional operations and artificial intelligence. It represents a shift from reactive troubleshooting to proactive, data-driven management of digital services. Unlike theoretical courses, this certification focuses on how to implement AI models within production-grade DevOps and SRE workflows.
It exists because modern enterprises require systems that can self-heal and predict failures before they impact customers. The program focuses on the practical application of algorithms to reduce “noise” in monitoring systems. It aligns with modern engineering practices by teaching professionals how to integrate automated insights directly into CI/CD pipelines and incident management platforms.
Who Should Pursue Certified AIOps Professional?
This certification is highly beneficial for DevOps engineers, Site Reliability Engineers (SREs), and Cloud Architects who are managing large-scale distributed systems. If you find yourself overwhelmed by alert fatigue or spending too much time on repetitive manual tasks, this track is specifically designed for your needs.
Security professionals and Data Engineers also benefit from this program, as it provides a framework for analyzing system behavior through a lens of anomaly detection. For engineering managers in India and across the globe, this certification provides the strategic knowledge needed to build and lead high-performing automation teams. Even beginners with a solid foundation in Linux and cloud can use this to future-proof their skills.
Why Certified AIOps Professional is Valuable and Beyond
The demand for AIOps skills is growing rapidly as companies realize that hiring more people is not a scalable solution for managing cloud-native complexity. Obtaining this certification ensures your longevity in the field by moving you from a “tool operator” to an “automation architect.” It helps you stay relevant even as specific tools like Jenkins or Ansible evolve or get replaced.
Enterprises are adopting AIOps to reduce Mean Time to Repair (MTTR) and improve service availability. By mastering these concepts, you become a high-value asset capable of delivering a significant return on investment for your organization. The career investment you make now will pay off as AI becomes a standard component of every enterprise IT stack.
Certified AIOps Professional Certification Overview
The program is delivered via the official portal at aiopsschool.com and is hosted on the AIOps School website. It features a tiered approach to learning, moving from foundational concepts to advanced architectural patterns. The assessment approach is designed to test your ability to solve real-world problems rather than just memorizing definitions.
The certification is structured to accommodate working professionals, offering flexible learning paths that include video tutorials, hands-on labs, and comprehensive documentation. Ownership of the certification resides with the hosting site, ensuring that the content is regularly updated to reflect the latest trends in the AIOps and MLOps ecosystems.
Certified AIOps Professional Certification Tracks & Levels
The certification is organized into three distinct levels: Foundation, Professional, and Advanced. The Foundation level introduces you to data collection and basic statistical analysis for IT operations. It is the starting point for anyone new to the integration of AI in infrastructure management.
The Professional level dives deeper into specific specialization tracks such as SRE-focused AIOps or FinOps-driven automation. Finally, the Advanced level is for those looking to design enterprise-wide AIOps strategies. These levels align with your career progression, moving from individual contributor roles to senior engineering and leadership positions.
Complete Certified AIOps Professional Certification Table
TrackLevelWho it’s forPrerequisitesSkills CoveredRecommended OrderOperationsFoundationJunior DevOps/SREBasic Linux & MonitoringData Ingestion, Pattern Matching1EngineeringProfessionalExperienced EngineersFoundation CertML Model Deployment, Root Cause Analysis2StrategyAdvancedArchitects/ManagersProfessional CertEnterprise AI Governance, Predictive Scaling3SpecializedSpecialistData EngineersPython BasicsAnomaly Detection, Log Clustering2 (Optional)OptimizationSpecialistFinOps PractitionersCloud Billing KnowledgeCost Prediction, Automated Right-sizing2 (Optional) Detailed Guide for Each Certified AIOps Professional Certification
Certified AIOps Professional – Foundation Level
What it is
This level validates your understanding of the core building blocks of AIOps, including data types and monitoring basics. It proves you can distinguish between traditional monitoring and AI-enhanced observability.
Who should take it
This is suitable for junior engineers, helpdesk professionals, or developers who want to understand how their applications behave in production. No prior AI experience is required.
Skills you’ll gain
Understanding the four stages of AIOps: Collection, Analysis, Action, and Reporting. Identifying different types of telemetry data: Logs, Metrics, Traces, and Events. Basic understanding of how machine learning differs from rule-based automation. Real-world projects you should be able to do
Set up a unified data dashboard that aggregates logs from multiple sources. Create a basic threshold-based alerting system that reduces redundant notifications. Preparation plan
7-14 Days: Focus on the official documentation and understanding the terminology. 30 Days: Complete the hands-on labs related to data ingestion and dashboarding. 60 Days: Review case studies of companies that successfully transitioned to AIOps. Common mistakes
Ignoring the importance of data quality before applying AI models. Confusing standard automation scripts with actual AIOps patterns. Best next certification after this
Same-track option: Certified AIOps Professional – Professional Level Cross-track option: SRE Practitioner Certification Leadership option: Technical Team Lead Workshop Certified AIOps Professional – Professional Level
What it is
This certification validates your ability to deploy and manage machine learning models specifically for IT operations. It focuses on solving complex problems like automated root cause analysis.
Who should take it
Experienced SREs, DevOps engineers, and Cloud Architects who have at least two years of experience in production environments.
Skills you’ll gain
Implementing supervised and unsupervised learning models for infrastructure. Automating the incident lifecycle using AI-driven insights. Configuring anomaly detection engines to find “silent” failures. Real-world projects you should be able to do
Build a predictive scaling system that adjusts resources based on traffic forecasts. Implement an automated incident correlation engine that groups related alerts. Preparation plan
7-14 Days: Deep dive into specific AIOps algorithms and their use cases. 30 Days: Work on model tuning and understanding precision vs. recall in IT data. 60 Days: Build a full-scale project demonstrating automated remediation. Common mistakes
Over-complicating models for simple problems that only need basic scripts. Failing to account for “model drift” in changing infrastructure environments. Best next certification after this
Same-track option: Certified AIOps Professional – Advanced Level Cross-track option: Certified MLOps Engineer Leadership option: Certified Engineering Manager Choose Your Learning Path
DevOps Path
The DevOps path focuses on integrating AI into the software development lifecycle. You will learn how to use AIOps to analyze build logs, predict deployment failures, and optimize CI/CD pipelines. This path is ideal for those who want to improve the velocity and reliability of software delivery through intelligent automation.
DevSecOps Path
In this path, the focus shifts to security-related automation. You will use AIOps concepts to identify unusual patterns that might indicate a security breach. It covers the use of AI for real-time threat detection and automated response to vulnerabilities within the infrastructure.
SRE Path
The SRE path is dedicated to maintaining high availability and performance. You will learn how to use AI to manage service level objectives (SLOs) and error budgets. This path emphasizes automated root cause analysis and the reduction of manual toil through smart incident management.
AIOps Path
This dedicated path focuses on the pure architecture of AI-driven operations. You will explore the deep technical layers of data pipelines and machine learning operations. It is designed for those who want to become specialists in building the actual platforms that power automated IT environments.
MLOps Path
The MLOps path bridges the gap between data science and production engineering. You will learn how to manage the lifecycle of machine learning models just like you manage code. This includes versioning models, monitoring their performance, and ensuring they are deployed reliably in cloud environments.
DataOps Path
DataOps focuses on the reliability and quality of the data that fuels AI models. You will learn how to build robust data pipelines that ensure information is accurate and available in real-time. This path is crucial for organizations that rely on large-scale data processing for their operational decisions.
FinOps Path
The FinOps path uses AI to solve the complex problem of cloud cost management. You will learn how to use predictive models to forecast cloud spending and identify waste. This path teaches you how to automate the right-sizing of resources to ensure maximum financial efficiency.
Role → Recommended Certified AIOps Professional Certifications
RoleRecommended CertificationsDevOps EngineerCertified AIOps Professional – Foundation & ProfessionalSRECertified AIOps Professional – Professional & AdvancedPlatform EngineerCertified AIOps Professional – Specialist (Anomaly Detection)Cloud EngineerCertified AIOps Professional – FoundationSecurity EngineerCertified AIOps Professional – Specialist (SecOps Focus)Data EngineerCertified AIOps Professional – DataOps SpecialistFinOps PractitionerCertified AIOps Professional – FinOps SpecialistEngineering ManagerCertified AIOps Professional – Advanced (Strategy) Next Certifications to Take After Certified AIOps Professional
Same Track Progression
Once you have mastered the professional level, you should look toward deep specialization. This involves taking advanced courses in specific areas like deep learning for log analysis or natural language processing for incident tickets. Staying within the same track allows you to become a recognized subject matter expert in the AIOps community.
Cross-Track Expansion
If you want to broaden your skill set, consider moving into related fields like MLOps or DevSecOps. Understanding how AI applies to security or how to manage ML models in production will make you a more versatile engineer. This cross-pollination of skills is highly valued in modern enterprise environments.
Leadership & Management Track
For those looking to transition away from hands-on coding, the leadership track is the logical next step. You can pursue certifications focused on engineering management and AI governance. This allows you to lead teams in implementing the complex strategies you learned during your technical certification journey.
Training & Certification Support Providers for Certified AIOps Professional
DevOpsSchool
DevOpsSchool is a leading platform that has established a strong reputation for providing high-quality technical training. They offer a comprehensive suite of resources for those looking to master AIOps and other modern engineering practices. Their approach combines theoretical knowledge with intense practical labs, ensuring that students can apply what they learn in real-world scenarios. With a large community of mentors and alumni, DevOpsSchool provides an environment where professionals can network and share experiences. Their curriculum is updated frequently to match the changing landscape of the IT industry, making them a reliable choice for long-term career development and certification support.
Cotocus
Cotocus specializes in delivering specialized training and consulting services for enterprise-level technologies. They focus on providing hands-on experience through industry-standard projects and real-world simulations. For the Certified AIOps Professional track, Cotocus offers tailored learning paths that help engineers understand the nuances of data-driven operations. Their trainers are experienced practitioners who bring insights from the trenches of software engineering and cloud management. Cotocus emphasizes the importance of mastering tools and methodologies that are currently in high demand. This makes them an excellent partner for both individual learners and corporate teams looking to upgrade their technical capabilities in the automation space.
Scmgalaxy
Scmgalaxy is a well-known resource hub and training provider that focuses on configuration management, DevOps, and automation. They provide a vast library of tutorials, guides, and certification prep materials that are highly beneficial for AIOps candidates. Scmgalaxy’s training programs are designed to be accessible yet deeply technical, covering everything from basic scripting to complex AI integrations. They have a strong presence in the Indian technical community and offer support for a wide range of certifications. Their focus on the practical aspects of software configuration and delivery makes them a go-to source for engineers looking to strengthen their foundational knowledge before moving into advanced AI topics.
BestDevOps
BestDevOps focuses on delivering premium training experiences that are centered on quality and career outcomes. They offer curated bootcamps and certification courses that are specifically designed to help professionals transition into senior roles. Their AIOps training modules are structured to be concise and impactful, focusing on the most relevant skills required by modern employers. BestDevOps prides itself on its hands-on approach, ensuring that every student completes significant project work during their course. By maintaining high standards for their instructors and course materials, they help ensure that their graduates are well-prepared for the challenges of managing production-grade AI-driven infrastructure.
devsecopsschool.com
This provider focuses specifically on the intersection of security and automation. They offer deep-dive training into how AI and machine learning can be used to secure modern cloud-native environments. For professionals taking the Certified AIOps Professional path, this site provides valuable context on anomaly detection and automated threat response. Their courses are designed for security engineers who want to adopt DevOps practices and for DevOps engineers who need to integrate security into their workflows. With a focus on practical security implementation, devsecopsschool.com is a vital resource for anyone looking to specialize in the security aspects of AIOps and intelligent automation.
sreschool.com
Sreschool.com is dedicated to the discipline of Site Reliability Engineering. They provide targeted training on maintaining system reliability, scalability, and performance using modern tools and techniques. Their AIOps-related content focuses on reducing toil and improving incident management through automated insights. Students can learn how to apply AI models to monitor service levels and manage error budgets effectively. The platform offers a variety of learning formats, including self-paced courses and instructor-led sessions. For SREs looking to add AIOps to their toolkit, sreschool.com provides the specific context and practical examples needed to succeed in a production environment.
aiopsschool.com
As the primary host for the Certified AIOps Professional program, aiopsschool.com is the most direct source of information and training for this certification. The site offers a comprehensive learning management system that tracks your progress through various levels of certification. They provide official study guides, practice exams, and access to specialized lab environments. The curriculum is developed by industry experts who are actively working in the AIOps field. Because they own the certification standards, their training materials are always perfectly aligned with the latest exam objectives. This makes aiopsschool.com the essential starting point for any professional pursuing this career path.
dataopsschool.com
Dataopsschool.com focuses on the management and orchestration of data pipelines, which is the foundation of any successful AIOps strategy. They teach professionals how to ensure data quality, accessibility, and security across the entire organization. For AIOps candidates, this provider offers crucial training on how to feed reliable data into machine learning models. Their courses cover data engineering, data governance, and the automation of data workflows. By mastering these skills at dataopsschool.com, engineers can ensure that their AIOps implementations are based on accurate and timely information, which is critical for making correct automated decisions in production.
finopsschool.com
Finopsschool.com addresses the financial management aspect of cloud operations. They provide training on how to use AI and data analysis to optimize cloud spending and improve financial accountability. For professionals in the Certified AIOps Professional program, this site offers specialized knowledge on cost-related anomaly detection and predictive billing. Their curriculum helps bridge the gap between finance and engineering teams, allowing for better collaboration on cloud efficiency. By learning the principles of FinOps through this provider, engineers can show tangible business value by automating the process of reducing waste and optimizing resource utilization in large-scale environments.
Frequently Asked Questions (General)
How difficult is the Certified AIOps Professional exam?
The exam is moderately challenging as it tests both theoretical concepts and practical application. If you have a solid background in operations and follow the recommended preparation plan, you will be well-positioned to pass. How long does it take to get certified?
Most professionals complete the training and pass the exam within 30 to 60 days. This depends on your prior experience with automation and the amount of time you can dedicate to studying each week. Are there any mandatory prerequisites for the Foundation level?
There are no mandatory prerequisites for the Foundation level, but a basic understanding of Linux, networking, and standard monitoring tools is highly recommended for a better learning experience. What is the return on investment for this certification?
The ROI is high, as AIOps specialists often command higher salaries compared to traditional DevOps engineers. It also opens doors to senior architect and leadership roles in top-tier technology companies. Do I need to be a data scientist to pass this?
No, you do not need to be a data scientist. The certification focuses on the application of AI models in an operational context, not on the deep mathematical creation of the models themselves. In what order should I take the certifications?
We recommend starting with the Foundation level, followed by the Professional level. Once those are secured, you can choose a specialist track or move to the Advanced level based on your career goals. Is the certification recognized globally?
Yes, the certification is recognized by enterprises and technology firms worldwide, including major markets in India, North America, and Europe. How often do I need to renew my certification?
The certification is typically valid for two to three years. Renewal usually involves taking an updated exam or completing continuing education credits through the platform. Does the program include hands-on labs?
Yes, the program emphasizes practical learning and includes various lab environments where you can practice data ingestion, model deployment, and incident automation. Can I take the exam online?
Yes, the exam is hosted online through the official website, allowing you to take it from the comfort of your home or office at a scheduled time. Are there community forums for students?
The program provides access to exclusive community forums where you can interact with other students, mentors, and industry experts to ask questions and share insights. What kind of support is available if I get stuck?
Students have access to technical support and mentorship through the training providers, ensuring that you can get help with complex topics or lab exercises. FAQs on Certified AIOps Professional
What specific AI models are covered in the curriculum?
The course covers a range of models including K-Means for log clustering, Random Forests for incident classification, and various time-series forecasting models for capacity planning and predictive scaling. How does this certification differ from a general AI course?
This program is strictly focused on IT operations. While a general AI course might focus on image recognition or natural language processing, this certification focuses on system telemetry, log patterns, and infrastructure health. Will this certification help me reduce manual work in my current job?
Absolutely. The core goal is to teach you how to automate repetitive tasks like alert sorting and root cause identification, allowing you to focus on more strategic engineering work. Do I need to know Python to succeed in this program?
While not strictly required for the Foundation level, a basic understanding of Python is very helpful for the Professional and Advanced levels, especially when working with data pipelines. Is AIOps relevant for small companies or only large enterprises?
While large enterprises benefit most due to their scale, small companies can use AIOps to manage complex cloud setups with a very small team, making it relevant for organizations of all sizes. Does the course cover tool-specific training like Splunk or Datadog?
The certification is designed to be vendor-neutral, focusing on the underlying principles and architectural patterns that can be applied to any modern monitoring or AIOps tool. Can this certification help me move into an SRE role?
Yes, AIOps is a core component of modern Site Reliability Engineering. This certification provides the automation skills that are essential for successful SRE practitioners. What is the focus of the Advanced level exam?
The Advanced level focuses on strategy, governance, and designing the overall AIOps architecture for an entire organization, including data privacy and model lifecycle management. Final Thoughts: Is Certified AIOps Professional Worth It?
As someone who has seen the evolution of IT from physical servers to complex cloud-native architectures, I can tell you that the era of manual monitoring is ending. The volume of data we generate today is simply too much for any human team to process effectively. Moving into AIOps is not just a career “upgrade”—it is a necessity for anyone who wants to remain competitive in the infrastructure space.
The Certified AIOps Professional program provides a structured, practical, and experience-driven way to gain these skills. It doesn’t just teach you how to use a tool; it teaches you how to think like an automation architect. If you are willing to put in the time to master these data-driven methodologies, you will find yourself at the forefront of the next major wave in enterprise technology. My advice is to start with the foundation and build your expertise steadily. The transition to intelligent operations is a journey, and this certification is the best roadmap you can have.
View the full article
iOS 27 is likely to introduce at least four new Apple Intelligence features that function within system apps, based on backend code discovered by Nicolás Alvarez and confirmed by MacRumors.


First up, Apple is expected to lean more heavily into Visual Intelligence in iOS 27, since the company is reportedly developing AI wearable devices that will leverage the feature. Apple is reportedly working on smart glasses, AirPods with cameras, and a wearable AI pin or pendant.

To that end, Visual Intelligence appears to be getting at least two new capabilities. One of them will likely let you scan a food nutrition label to get more information, which could well integrate into the Health app. Another will offer to add printed phone numbers/addresses to your Contacts. Visual Intelligence already offers to add calendar dates to your Calendar, so an equivalent feature for contacts makes sense.

Elsewhere, Apple's Wallet app is likely to gain the ability to generate digital passes from scans of things like event tickets, gym membership cards and the like. Google Wallet for Android already does something similar, using AI to determine the content of a pass.

Meanwhile, in Safari, we're expecting a new AI feature that will automatically name Tab Groups for users based on the contents of the tabs within the group.

We can't say with certainty that the above features will work as described, since we're interpreting them from the names of individual code strings. Likewise, we do not know for sure that they will appear in iOS 27 or a future point update of the upcoming software, but given that Apple is working on a smarter version of ‌Siri‌ for iOS 27 with deeper integration across apps, the timing fits.

Apple will unveil iOS 27 at the Worldwide Developers Conference in June, before launching in September just ahead of when new iPhone models come out.Related Roundup: iOS 27Tags: Apple Intelligence, Visual Intelligence
This article, "4 New Apple Intelligence Features Found in Apple Code, Likely in iOS 27" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
A "novel" social engineering campaign has been observed abusing Obsidian, a cross-platform note-taking application, as an initial access vector to distribute a previously undocumented Windows remote access trojan called PHANTOMPULSE in attacks targeting individuals in the financial and cryptocurrency sectors. Dubbed REF6598 by Elastic Security Labs, the activity has been found to leverageView the full article
A bank approved a Taboola pixel. That pixel quietly redirected logged-in users to a Temu tracking endpoint. This occurred without the bank’s knowledge, without user consent, and without a single security control registering a violation. Read the full technical breakdown in the Security Intelligence Brief. Download now → The "First-Hop Bias" Blind Spot Most&View the full article
Several major insurance carriers have begun to back away from providing cybersecurity and other insurance to companies using AI to run internal processes, insiders say.
While there’s no standard response to customer use of AI in the insurance market, many carriers are now quietly declining to write policies for claims related to AI-generated outputs in cybersecurity and errors and omissions (E&O) coverage, these observers say. Other insurance carriers are jacking up prices to cover AI-related claims, they say.
Dozens of insurance carriers appear to be rethinking coverage for mistakes related to AI, says Connor Deeks, CEO of Codestrap, an AI development and consulting firm that works with insurance firms.
Many insurance companies aren’t comfortable with covering AI outputs because they can’t track the reasoning path the AI took to come up with a result, he says.
“That’s playing out downstream with insurance companies basically carving out coverage, whether that’s across cybersecurity or E&O,” he says. “All of these vibe-coded solutions and these AI systems that people have constructed have inherent risk baked into the cake now, and you can’t actually see the full process.”
The insurance carrier concerns about AI workloads first surfaced in November 2025, when Financial Times reported that three major carriers, AIG, Great American, and W.R. Berkley, filed requests with US regulators to offer insurance policies that exclude liabilities tied to AI tools such as chatbots and agents. At the time, those requests appeared to be preemptive moves to be allowed to exclude AI mistakes sometime in the future.
But now, many carriers seem to be moving forward with plans to exclude AI mistakes from policies, Deeks says. Several carriers he’s been in contact with are moving to limit or end coverage for AI-related business disruptions and liabilities, he adds. The irony is that many insurance carriers are embracing AI for their own internal purposes.
Deeks’ company has a vested interest in AI insurance coverage — Codestrap markets its AI coding platform as traceable and therefore insurable — but other industry insiders have also seen similar carrier decisions.
Carriers find exclusions
It’s still unclear how many carriers will refuse to insure AI workloads, but several carriers are now writing insurance policies that exempt coverage for AI-related business chaos, says Jason Bishara, financial practice leader at global carrier NSI Insurance Group.
“The risk appetite is changing among the carriers, and it’s always constantly evolving,” he says. “With regard to AI, there are carriers that are just removing it from their risk appetite and declining to quote altogether.”
While some carriers have declined to cover AI outputs, others are building in rate hikes to cover the increased risk, Bishara says. While he doesn’t have numbers on the extent of the rate hikes, they are significant, he adds.
“Every business has insurance, and every business now is using AI to some extent,” he adds. “Are you seeing those liabilities and exclusions within these policies and an aversion to it from the carriers? The answer is yes.”
Carriers are also treating AI vendors differently than AI users, he says. In many cases, carriers are declining to cover AI vendors altogether, while they carve out exceptions in policies against covering AI at companies using the technology.
“If you’re an AI-related company or specifically an AI company, there’s a good chance that you’ll get a declination at this point,” he adds.
In recent months, many carriers have been asking detailed questions about how customers are using AI to better understand the risk of insuring potential mishaps, he says. Ultimately, this increased scrutiny will make it more difficult for companies to buy insurance for AI workloads.
“For everybody leveraging AI right now, you’re seeing questions like, ‘What are your AI policies? What are your procedures? How are you leveraging AI within your business?’” Bishara adds. “We’re getting a lot of questions from the underwriters on, ‘How do you leverage AI within your business?’”
Coverage in flux
Phil Karecki, CTO for the insurance sector at managed services provider Ensono, also sees some carriers backing away from covering AI outputs, although he’s not sure whether it’s a major trend. Insurance carriers continuously experiment with how to provide coverage, he notes.
Carriers have tried to separate tightly governed AI deployments from more experimental projects when determining whether to provide coverage, he says.
“You’ve got this bifurcation of AI, the governed generative and the autonomous pieces,” he says. “It’s no longer, ‘Are you using AI?’ It’s asking, ‘Are you using governed AI? How are you governing it? How are you keeping it safe and secure?’”
Carriers have been trying to determine whether covering AI workloads can be profitable for them, Karecki adds. Governed AI tools operating in a bounded decision-making process will be more insurable, while experimental AI systems with no monitoring and no easy rollback will be difficult to cover, he notes.
“There’s a repositioning versus a pullback, and that’s very common to the industry, and they will at times open up coverage just to see if it’s this type of insurance that will sell,” he says. “They will assess the results and what needs to change so they can decide whether to re-enter this marketplace or abandon it completely.”
In some cases, whether an AI system is insurable may come down to circumstances at individual insurance customers. Carriers in general don’t want to get out of the business of providing insurance, Karecki says.
“What they’re working for right now is, ‘How do I make this profitable, and is this sector insurable?’” he says. “They make those decisions on every application regardless, but now, depending upon what they’re being asked to insure, the questions will follow. ‘What are you using AI for? How are you governing it? What risks does that introduce?’”
It makes sense that some carriers have begun to question whether to cover AI outputs, given the current level of unreliability of most AI systems, says Dorian Smiley, CTO at Codestrap.
“The math says these models should be deterministic, like given the same input, you should get the same output,” he says. “But you can get very different output from the same input, and they can’t know if the answer that they’re giving you is actually correct.”
In most cases, AI models lack inductive reason and can’t review their own work, but many organizations are talking about deploying hundreds of autonomous agents and treating them like digital employees, he notes.
“The idea that these agents are going to become employees, autonomous people working in your organization, is insane,” he says. “You would never hire a person that can’t learn new information, can’t reliably retrieve information, or check their own work.”
NSI’s Bishara has advice for IT and business leaders looking for insurance coverage for their AI workloads: Be honest about how they’re using AI. If they try to hide their AI risks, they risk having their claims rejected when something goes wrong, he says.
“If you don’t fully disclose these things appropriately in the way in which you’re functioning and operating, it could be utilized as an excuse to deny a claim at a later date,” he says. “You don’t want a carrier to come back and say, ‘We didn’t underwrite to that risk. We asked these questions, and you didn’t disclose it.’”
View the full article
Apple has started ramping up its supply chain for a new variable aperture camera system expected to debut in the iPhone 18 Pro and Pro Max this September, reports Korea's ETNews.


Apple has never implemented a variable aperture on an iPhone. From the iPhone 14 Pro through the iPhone 17 Pro, the main camera uses a fixed ƒ/1.78 aperture, meaning the lens remains fully open at all times when capturing images.

In contrast, a variable aperture lets the camera control how much light reaches the sensor. In low-light conditions, it opens to admit more light, while in bright scenes, it closes to avoid overexposure. This should also give users more control over depth of field.

In December 2024, Apple supply chain analyst Ming-Chi Kuo was first to say that that the main rear camera on both ‌iPhone 18 Pro‌ models will offer variable aperture. A report from October 2025 said Apple was moving ahead with plans to bring the technology to next-generation iPhones and was discussing components with suppliers.

According to industry sources cited by today's report, Chinese supplier Sunny Optical has already started producing the actuators that enable the aperture mechanism, while camera module assembly will follow in early summer. Apple's primary camera partner LG Innotek is said to be preparing to begin production around June or July, with dedicated equipment being installed at its Gumi facility in South Korea. Module makers such as Cowell are also expected to take part in the process.

LG Innotek is reportedly likely to take on a larger share of production for the main camera module due to the system's added complexity. A similar thing happened when Apple adopted a folded zoom lens in the iPhone 15 Pro Max, where LG Innotek initially served as the sole supplier.
10 Reasons to Wait for the iPhone 18 Pro
The ‌iPhone 18 Pro‌ and ‌iPhone 18 Pro‌ Max are expected to be announced alongside Apple's first foldable iPhone sometime around September.Related Roundup: iPhone 18 Pro
This article, "iPhone 18 Pro Variable Aperture Camera Enters Production" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
It is difficult to understand why, in 2026, we are still debating the reporting line of the chief information security officer (CISO).
It is one of the first topics I wrote about in 2015, and after more than two decades of high-profile cyber incidents, sustained regulatory pressure, massive technology investments and the steady elevation of cybersecurity to boardroom agendas, one might reasonably expect that this issue would have been settled long ago.
Yet the question persists. And articles like this It’s time to rethink CISO reporting lines show that the debate is still raw.
The fact that the debate continues tells us something important. It reveals that many organizations still struggle with a more fundamental question: What exactly is the role of the CISO within the enterprise?
The reporting line matters — but it was never the real question
Let me be clear. The reporting line matters. It matters because it defines the authority, visibility and influence of the security function across the organization. It signals internally how seriously cybersecurity is taken and determines how effectively the CISO can engage with the executive leadership team.
But the reporting line was never the real question.
The real question is whether the CISO has the organizational standing necessary to influence decisions across multiple silos: IT, operations, legal, compliance, HR, procurement, third-party suppliers and increasingly a complex ecosystem of partners and digital platforms.
Cybersecurity is one of the very few corporate functions that touch virtually every part of the enterprise. It is therefore inherently cross-functional. Without sufficient authority and visibility, the CISO cannot hope to influence behaviour across the organization, let alone drive meaningful change.
If we are still debating the reporting line in 2026, it is largely because many organizations still treat cybersecurity as a technical issue rather than a leadership issue.
The governance gap behind the debate
The persistence of this debate reflects a broader governance gap.
Historically, information security emerged as a technical discipline embedded within IT departments. Early security teams focused primarily on protecting infrastructure: Firewalls, access controls, network monitoring and vulnerability management. In that environment, it was natural for the security function to sit within the IT organization.
But the nature of cyber risk has evolved dramatically.
Cybersecurity today is not merely about protecting technology infrastructure. It is about protecting digital business models, customer trust, intellectual property, operational resilience and in some sectors even national security interests.
In other words, cybersecurity has become a strategic business issue.
And yet, in many organizations, the governance structures surrounding cybersecurity have not evolved at the same pace.
The continuing debate about the CISO reporting line is therefore less about organizational design and more about whether companies have fully internalised the strategic nature of cyber risk.
There is no universal reporting line
Another recurring misconception is the search for a universal answer.
Every year, surveys attempt to determine the “correct” reporting line for the CISO. Some conclude that the CISO should report to the CEO. Others recommend the CRO or the COO. Some insist that independence from IT is essential.
In reality, there is no universal model. The reporting line remains a means to an end.
Organizations differ widely in their structure, culture, maturity and regulatory environment. What works in one organization may not work in another.
In many organizations, the CIO remains the most natural reporting line for the CISO, particularly where technology transformation and digital innovation are core strategic priorities. In others, the COO or the CEO may be better placed to support the operational changes required to embed security across business processes.
What matters is not the job title of the executive above the CISO.
What matters is whether that individual has the authority, credibility, organizational reach and personal willingness to support the security agenda.
Authority matters — and quite a lot of that is forged in the first 100 days
When a new CISO joins an organization, their immediate priority is rarely technical. Instead, it is organizational: Understanding the business, mapping stakeholders, assessing governance structures and identifying the cultural barriers that may hinder security improvements.
During those first months, the CISO must build credibility quickly across multiple constituencies. They must engage with senior executives, operational leaders, technology teams and sometimes regulators or external partners.
None of this can be done effectively if the CISO lacks organizational authority.
A reporting line that leaves the CISO buried several layers below executive leadership severely limits their ability to build the relationships required to succeed. Conversely, a reporting line that provides direct access to senior decision-makers can dramatically accelerate the process.
The reporting line, therefore, matters not because it determines technical decisions, but because it determines access, influence and credibility.
The illusion of structural solutions
At the same time, we should be careful not to overstate the importance of organizational charts.
A common mistake is to assume that moving the CISO reporting line will automatically solve cybersecurity challenges.
It will not.
Cybersecurity failures rarely occur because the organizational chart was incorrect. They occur because of poor governance, weak leadership, unclear accountability or cultural resistance to change.
The most effective CISOs succeed not because of perfect reporting structures but because they build trust, credibility and influence across the organization.
Which brings us to perhaps the most important factor of all: The relationship between the CISO and their direct superior.
Trust matters more than structure
In practice, the success of the CISO depends heavily on the quality of the relationship with the executive to whom they report.
That relationship must be built on trust, alignment and shared understanding of the organization’s risk appetite and strategic priorities.
If the executive above the CISO understands the importance of cybersecurity and is willing to champion the security agenda at the board level and across the firm, the reporting structure can work extremely well.
If that support is absent because the business at large does not see the strategic importance of cybersecurity, no reporting line will magically solve the problem.
The myth of the CIO–CISO conflict
One final argument frequently raised in these discussions is the supposed “conflict of interest” between the CIO and the CISO.
According to this theory, the CISO should not report to the CIO because the CIO is responsible for delivering technology projects and operational performance, while the CISO is responsible for enforcing security controls that may slow things down.
This argument may have had some relevance 20 years ago, when security functions were primarily responsible for auditing IT operations.
But today, it increasingly reflects an outdated understanding of both roles.
Modern cybersecurity is deeply intertwined with technology architecture, cloud platforms, DevOps pipelines, digital transformation programs and operational resilience initiatives. Security cannot be treated as an external oversight function policing IT from a distance.
It must be embedded within technology strategy itself. Any modern CIO should see it that way.
In that environment, close collaboration between the CIO and the CISO is not only desirable — it is essential.
Framing the relationship as a structural budgetary conflict and a source of friction is counterproductive and outdated. The real objective should not be to avoid friction but to engineer alignment: Ensuring that technology leadership and security leadership work together to support the organization’s strategic goals.
Moving beyond the debate
Ultimately, the continuing debate about the CISO reporting line distracts the security industry from more important questions.
What matters far more is whether cybersecurity is integrated into corporate governance, supported by executive leadership and aligned with business strategy.
If organizations are still arguing about where the CISO should sit in 2026, it may simply indicate that they have not yet fully accepted the strategic nature of cyber risk.
And until that changes, the debate will likely continue.
Not because the answer is difficult — but because the underlying governance challenge remains unresolved.
This article is published as part of the Foundry Expert Contributor Network.
Want to join?
View the full article
Name : CyberSec India Expo 2026
CyberSec India Expo 2026 l Where Cybersecurity Strategy Meets Real-World Impact
Date: April 23-24, 2026 Venue: Bombay Exhibition Centre, Mumba
The CyberSec India Expo 2026, now in its 2nd edition, offers the entire cybersecurity ecosystem a comprehensive platform to showcase cutting-edge solutions driving India’s cybersecurity vision.
Following a highly successful inaugural edition in 2025—welcoming 3,400+ attendees, 45+ expert speakers, and enabling meaningful business interactions with 65+ exhibitors and leading brands—the 2026 edition promises to be bigger, broader, and more impactful.
This year’s Expo will feature an expanded showcase of next-generation cybersecurity technologies alongside a power-packed conference centered on the theme: “Advancing India’s National Cyber Resilience in a Digital-First Economy – Securing Digital India with AI-enabled defense, Zero Trust, and future-ready quantum security.”
Website : https://www.cybersecindiaexpo.com/
Free Exhibition Registration : https://bit.ly/4uSPIEz
Conference Registration : https://bit.ly/4rRvxE7
Coupon code for discount : Ec3
The post CyberSec India Expo 2026 appeared first on CISO MAG | Cyber Security Magazine.
View the full article
Name : CISO India Connect 2026 – Bengaluru
“CISO INDIA CONNECT 2026 arrives in Bangalore on 16th April, bringing together leading Chief Information Security Officers, risk leaders, and cybersecurity innovators for a focused day of strategic dialogue and collaboration. Hosted in association with ISC2 and the ISACA Bangalore Chapter, the event serves as a premier platform for knowledge exchange and executive networking within India’s dynamic security community.
The conference will feature expert led sessions on emerging threat landscapes, governance and compliance strategies, cloud security, AI driven defense, and board level cyber risk management. Attendees will gain practical insights from industry veterans, participate in peer discussions, and explore innovative security solutions from leading technology partners.
CISO India Connect continues its mission of strengthening the cybersecurity ecosystem by fostering meaningful partnerships, advancing professional excellence, and enabling security leaders to navigate evolving digital risks with confidence and resilience.”
Registration link: https://form.jotform.com/260332089606456
The post CISO India Connect 2026 – Bengaluru appeared first on CISO MAG | Cyber Security Magazine.
View the full article
The Computer Emergencies Response Team of Ukraine (CERT-UA) has disclosed details of a new campaign that has targeted governments and municipal healthcare institutions, mainly clinics and emergency hospitals, to deliver malware capable of stealing sensitive data from Chromium-based web browsers and WhatsApp. The activity, which was observed between March and AprilView the full article
DC Studio | shutterstock.com
Konfigurationsfehler in der Cloud, die Unternehmensdaten gefährden, sind nicht unbedingt etwas Neues – eher im Gegenteil. Umso schlimmer, dass Unternehmen ihre Cloud-Ressourcen immer noch nicht durchgängig absichern. Zumindest legt das ein aktueller Report nahe. Dafür hat der Cloud-Sicherheitsanbieter Qualys 101 Cybersecurity- und IT-Profis befragt, zu deren Aufgaben es gehört, Cloud-Umgebungen abzusichern. Demnach:
haben 28 Prozent der Befragten im vergangenen Jahr einen Breach in Zusammenhang mit der Cloud oder SaaS-Applikationen verzeichnet. sehen 24 Prozent falsch konfigurierte Services als das größte Risiko für ihre Cloud-Umgebung an. Qualys nahm für seine Studie außerdem rund 44 Millionen virtuelle Maschinen (VMs) unter die Lupe, die in Public Clouds gehostet werden. Dabei stellten die Experten fest, dass 45 Prozent der AWS-VMs, 63 Prozent der GCP-VMs und 70 Prozent der Azure-VMs über falsch konfigurierte Ressourcen verfügten.
Die häufigsten Cloud-Konfigurationsfehler
Laut Ayan Roy, Leiter des Bereichs Cybersicherheit bei EY Americas, aktivieren Unternehmen zwar durchaus bestimmte Cloud-Security-Funktionen, allerdings nicht alle. So blieben Logging, Monitoring und Multi-Faktor-Authentifizierung (MFA) in vielen Fällen unbeachtet: “Unternehmen wollen schnell vorankommen, und die Time-to-Value ist absolut entscheidend. Wenn Cybersicherheitsteams jedoch nicht in diese Entscheidungen eingebunden werden, beginnen die Probleme. So können die Sicherheitsprofis oft nur nachträglich Maßnahmen ergreifen.”
Einen weiteren blinden Fleck in Sachen Cloud Security sieht Roy während Fusionen und Übernahmen. Er mahnt Unternehmen dazu, in solchen Fällen proaktiv vorzugehen: “Führen Sie eine Due Diligence durch, berücksichtigen Sie diese auch und stellen Sie sicher, dass Sie den richtigen Cybersecurity-Investitionsplan haben.”
Laut Scott Wheeler, Cloud Practice Lead bei Asperitas, treten mit steigender Unternehmensgröße auch weniger Cloud-Konfigurationsfehler auf. Das liegt laut dem Cloud-Experten vor allem an der Aufsicht durch Regulierungsbehörden. Kleine Firmen hätten hingegen enorme Probleme, da sie weder über das Personal noch die nötigen Tools verfügten, um Risiken im Zusammenhang mit der Cloud-Konfiguration zu managen – etwa exponierte Speicher-Buckets oder Web Services.  
“Das gesamte Konzept von Zero Trust basiert auf der Tatsache, dass man den Zugriff auf das benötigte Minimum beschränken kann. Aber das ist in der Praxis schwer zu bewerkstelligen”, erklärt Wheeler. Oftmals würden während der Entwicklung Berechtigungen erweitert und nach der Inbetriebnahme nicht wieder zurückgesetzt, wie er beispielhaft anführt. Der regelmäßig größte Fehler, den Wheeler beobachtet, sind jedoch Datenbanken oder andere Cloud-Assets, die über nicht sichere, private Netzwerke kommunizieren. “Viele dieser Services unterstützen das nicht ‘out of the box’. Es erfordert einiges an Arbeit, sie so zu konfigurieren, dass ausschließlich privater Netzwerkverkehr in private Cloud- oder lokale Umgebungen fließt. Das ist ein großes Problem, das oft von kriminellen Hackern ausgenutzt wird.”
Und auch wenn viele Anbieter versprechen, dass KI auch Cloud Security einfacher, kostengünstiger und effektiver gestalten wird: Sie sollten nicht damit rechnen, dass Cloud-Konfigurationsprobleme schon morgen der Vergangenheit angehören. Bis KI-Agenten soweit sind, dass sie das zuverlässig übernehmen können, wird noch ein bisschen Zeit ins Land ziehen.
9 Tipps für sicherere Cloud-Konfigurationen
Tun können Sie dennoch etwas gegen fehlerhafte Cloud-Konfigurationen. Zum Beispiel:
1. Multi-Faktor-Authentifizierung implementieren
MFA sollte für jede Form von Cloud-Zugriff zur Anwendung kommen, nicht nur für bestimmte Benutzer.
2. Private Netzwerke für alle Services nutzen
Konfigurieren Sie Datenbanken und Cloud-Dienste so, dass sie nur über private Netzwerke und nicht über das öffentliche Internet kommunizieren.
3. Daten verschlüsseln
Datenverschlüsselung sollte für alle neuen und bestehenden Ressourcen standardmäßig aktiviert sein. Angesichts des nahenden Quanten-Zeitalters empfiehlt es sich für Unternehmen bereits jetzt auf quantensichere Verschlüsselungsalgorithmen zu setzen, um sich gegen sogenannte “Harvest now, decrypt later”-Angriffe zu schützen.
4. Least-Privilege-Zugriffskontrollen anwenden
Benutzern und Systemen Zugriff auf möglichst wenige Ressourcen zu gewähren, ist ein Grundpfeiler moderner Zero-Trust-Sicherheitsprinzipien. Konten mit übermäßigen Berechtigungen können schnell zu Datenverlust führen, wenn sie missbraucht werden.
5. Infrastructure as Code verwenden
Wenn Administratoren oder Benutzer Änderungen an Cloud-Konfigurationen in den Cloud-Management-Konsolen vornehmen, ist es oft schwierig, diese nachzuvollziehen – und rückgängig zu machen, wenn etwas schiefgeht. Das Prinzip von Infrastructure as Code hilft an dieser Stelle: Verwenden Sie entsprechende Konfigurationsmanagement-Tools, um sämtliche Änderungen anhand von Richtlinien zu überprüfen, nachzuverfolgen und auditieren zu können.
6. Kontinuierlich scannen
Einmal bei der Ersteinrichtung der Cloud-Ressourcen zu überprüfen, ob die Konfigurationen aktuell sind, reicht nicht aus. Unternehmen müssen sicherstellen, dass sich nichts verändert. Zu diesem Zweck haben einige Cloud-Anbieter native Tools im Angebot. Lösungen aus dem Bereich Cloud Security Posture Management (CSPM) können außerdem dabei unterstützen, Lücken zu schließen oder Multi-Cloud-Umgebungen zu überwachen.
7. Speicher-Buckets sperren
Unsichere Amazon-S3-Buckets waren vor einigen Jahren bei Cyberkriminellen sehr populär – und sind nach wie vor ein gängiges Problem für Unternehmen. Um sicherzustellen, dass der Storage standardmäßig privat ist, empfehlen sich Bucket Policies und -Zugriffskontrollen.
8. Logging und Monitoring einziehen
Viele Unternehmen überwachen essenzielle Cloud Services, dabei bleibt Schatten-IT jedoch oft außen vor. Das ist weniger ein technologisches, als vielmehr ein Management-Problem und lässt sich durch bessere Kommunikation mit den Geschäftsbereichen und einen disziplinierteren Ansatz bei der Bereitstellung von Technologien lösen.
9. Security by Design verinnerlichen
Integrieren Sie Sicherheit von Anfang an in Ihre Cloud-Architektur – es ist immer deutlich schwieriger, nachträglich aufzurüsten. (fm)
View the full article
DC Studio | shutterstock.com
Konfigurationsfehler in der Cloud, die Unternehmensdaten gefährden, sind nicht unbedingt etwas Neues – eher im Gegenteil. Umso schlimmer, dass Unternehmen ihre Cloud-Ressourcen immer noch nicht durchgängig absichern. Zumindest legt das ein aktueller Report nahe. Dafür hat der Cloud-Sicherheitsanbieter Qualys 101 Cybersecurity- und IT-Profis befragt, zu deren Aufgaben es gehört, Cloud-Umgebungen abzusichern. Demnach:
haben 28 Prozent der Befragten im vergangenen Jahr einen Breach in Zusammenhang mit der Cloud oder SaaS-Applikationen verzeichnet. sehen 24 Prozent falsch konfigurierte Services als das größte Risiko für ihre Cloud-Umgebung an. Qualys nahm für seine Studie außerdem rund 44 Millionen virtuelle Maschinen (VMs) unter die Lupe, die in Public Clouds gehostet werden. Dabei stellten die Experten fest, dass 45 Prozent der AWS-VMs, 63 Prozent der GCP-VMs und 70 Prozent der Azure-VMs über falsch konfigurierte Ressourcen verfügten.
Die häufigsten Cloud-Konfigurationsfehler
Laut Ayan Roy, Leiter des Bereichs Cybersicherheit bei EY Americas, aktivieren Unternehmen zwar durchaus bestimmte Cloud-Security-Funktionen, allerdings nicht alle. So blieben Logging, Monitoring und Multi-Faktor-Authentifizierung (MFA) in vielen Fällen unbeachtet: “Unternehmen wollen schnell vorankommen, und die Time-to-Value ist absolut entscheidend. Wenn Cybersicherheitsteams jedoch nicht in diese Entscheidungen eingebunden werden, beginnen die Probleme. So können die Sicherheitsprofis oft nur nachträglich Maßnahmen ergreifen.”
Einen weiteren blinden Fleck in Sachen Cloud Security sieht Roy während Fusionen und Übernahmen. Er mahnt Unternehmen dazu, in solchen Fällen proaktiv vorzugehen: “Führen Sie eine Due Diligence durch, berücksichtigen Sie diese auch und stellen Sie sicher, dass Sie den richtigen Cybersecurity-Investitionsplan haben.”
Laut Scott Wheeler, Cloud Practice Lead bei Asperitas, treten mit steigender Unternehmensgröße auch weniger Cloud-Konfigurationsfehler auf. Das liegt laut dem Cloud-Experten vor allem an der Aufsicht durch Regulierungsbehörden. Kleine Firmen hätten hingegen enorme Probleme, da sie weder über das Personal noch die nötigen Tools verfügten, um Risiken im Zusammenhang mit der Cloud-Konfiguration zu managen – etwa exponierte Speicher-Buckets oder Web Services.  
“Das gesamte Konzept von Zero Trust basiert auf der Tatsache, dass man den Zugriff auf das benötigte Minimum beschränken kann. Aber das ist in der Praxis schwer zu bewerkstelligen”, erklärt Wheeler. Oftmals würden während der Entwicklung Berechtigungen erweitert und nach der Inbetriebnahme nicht wieder zurückgesetzt, wie er beispielhaft anführt. Der regelmäßig größte Fehler, den Wheeler beobachtet, sind jedoch Datenbanken oder andere Cloud-Assets, die über nicht sichere, private Netzwerke kommunizieren. “Viele dieser Services unterstützen das nicht ‘out of the box’. Es erfordert einiges an Arbeit, sie so zu konfigurieren, dass ausschließlich privater Netzwerkverkehr in private Cloud- oder lokale Umgebungen fließt. Das ist ein großes Problem, das oft von kriminellen Hackern ausgenutzt wird.”
Und auch wenn viele Anbieter versprechen, dass KI auch Cloud Security einfacher, kostengünstiger und effektiver gestalten wird: Sie sollten nicht damit rechnen, dass Cloud-Konfigurationsprobleme schon morgen der Vergangenheit angehören. Bis KI-Agenten soweit sind, dass sie das zuverlässig übernehmen können, wird noch ein bisschen Zeit ins Land ziehen.
9 Tipps für sicherere Cloud-Konfigurationen
Tun können Sie dennoch etwas gegen fehlerhafte Cloud-Konfigurationen. Zum Beispiel:
1. Multi-Faktor-Authentifizierung implementieren
MFA sollte für jede Form von Cloud-Zugriff zur Anwendung kommen, nicht nur für bestimmte Benutzer.
2. Private Netzwerke für alle Services nutzen
Konfigurieren Sie Datenbanken und Cloud-Dienste so, dass sie nur über private Netzwerke und nicht über das öffentliche Internet kommunizieren.
3. Daten verschlüsseln
Datenverschlüsselung sollte für alle neuen und bestehenden Ressourcen standardmäßig aktiviert sein. Angesichts des nahenden Quanten-Zeitalters empfiehlt es sich für Unternehmen bereits jetzt auf quantensichere Verschlüsselungsalgorithmen zu setzen, um sich gegen sogenannte “Harvest now, decrypt later”-Angriffe zu schützen.
4. Least-Privilege-Zugriffskontrollen anwenden
Benutzern und Systemen Zugriff auf möglichst wenige Ressourcen zu gewähren, ist ein Grundpfeiler moderner Zero-Trust-Sicherheitsprinzipien. Konten mit übermäßigen Berechtigungen können schnell zu Datenverlust führen, wenn sie missbraucht werden.
5. Infrastructure as Code verwenden
Wenn Administratoren oder Benutzer Änderungen an Cloud-Konfigurationen in den Cloud-Management-Konsolen vornehmen, ist es oft schwierig, diese nachzuvollziehen – und rückgängig zu machen, wenn etwas schiefgeht. Das Prinzip von Infrastructure as Code hilft an dieser Stelle: Verwenden Sie entsprechende Konfigurationsmanagement-Tools, um sämtliche Änderungen anhand von Richtlinien zu überprüfen, nachzuverfolgen und auditieren zu können.
6. Kontinuierlich scannen
Einmal bei der Ersteinrichtung der Cloud-Ressourcen zu überprüfen, ob die Konfigurationen aktuell sind, reicht nicht aus. Unternehmen müssen sicherstellen, dass sich nichts verändert. Zu diesem Zweck haben einige Cloud-Anbieter native Tools im Angebot. Lösungen aus dem Bereich Cloud Security Posture Management (CSPM) können außerdem dabei unterstützen, Lücken zu schließen oder Multi-Cloud-Umgebungen zu überwachen.
7. Speicher-Buckets sperren
Unsichere Amazon-S3-Buckets waren vor einigen Jahren bei Cyberkriminellen sehr populär – und sind nach wie vor ein gängiges Problem für Unternehmen. Um sicherzustellen, dass der Storage standardmäßig privat ist, empfehlen sich Bucket Policies und -Zugriffskontrollen.
8. Logging und Monitoring einziehen
Viele Unternehmen überwachen essenzielle Cloud Services, dabei bleibt Schatten-IT jedoch oft außen vor. Das ist weniger ein technologisches, als vielmehr ein Management-Problem und lässt sich durch bessere Kommunikation mit den Geschäftsbereichen und einen disziplinierteren Ansatz bei der Bereitstellung von Technologien lösen.
9. Security by Design verinnerlichen
Integrieren Sie Sicherheit von Anfang an in Ihre Cloud-Architektur – es ist immer deutlich schwieriger, nachträglich aufzurüsten. (fm)
View the full article
Apple retail locations and Apple Authorized Service Providers will soon be able to restore Apple Watch software in-store without needing to send an Apple Watch to a service center, according to a retail source that spoke to MacRumors.


Right now, Apple Watches that can't be restored using an iPhone need to be mailed to an Apple Repair Center for service. There is no in-store repair option, so customers have to wait for the Apple Watch to be shipped to the repair depot, get repaired, and be shipped back.

Starting later this month, Apple Stores and AASPs will be able to use an Apple Watch repair dock that connects to a Mac to restore the software on an Apple Watch. An in-store option for fixing software will make software-based repairs much quicker.

With watchOS 8.5 and iOS 15.4, Apple introduced an iPhone-based wireless restore option, but it is limited. It can only be used when a restore prompt is shown on the Apple Watch. For software issues where the iPhone restore doesn't work, the Apple Watch needs a specialized repair currently unavailable in retail stores. Failed updates, bricked devices, and boot loops can't be fixed with an iPhone.

Early Apple Watch models had a diagnostic port that Apple Stores could use for software fixes, but it was removed with the Apple Watch Series 7, and Apple switched to a wireless restoration process. After the port was dropped, Apple Watch software repairs had to be done at Apple Service Centers, making software-based failures a hassle for customers.Related Roundup: Apple Watch 11Buyer's Guide: Apple Watch (Neutral)
This article, "Apple Stores Will Soon Be Able to Restore Apple Watch Software In-House" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Apple today stopped signing iOS 26.4, so iPhone users who have updated to iOS 26.4.1 are no longer able to downgrade to the earlier version of iOS. iOS 26.4.1 came out a week ago.


When software is "signed," it means it can pass the server-side verification check that Apple does when a user downloads a new version of iOS on an iPhone. An update that's not signed can't be installed because it won't pass the verification check.

Apple does not show users earlier versions of iOS after an upgrade has been released, but when software is still signed, it is possible to downgrade with the macOS Finder app on a Mac or the Apple Devices app on a Windows PC.

Unsigning software prevents Apple customers from installing outdated, less secure versions of iOS, and Apple typically stops signing an update a week or so after new software comes out.

iOS 26.4.1 fixed iCloud syncing issues and Stolen Device Protection on enterprise devices. While iOS 26.4.1 is the current publicly available version of iOS, Apple is also beta testing an iOS 26.5 update.Related Roundups: iOS 26, iPadOS 26Related Forum: iOS 26
This article, "iOS 26.4 No Longer Signed by Apple, Blocking Downgrades From iOS 26.4.1" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
An iPhone exploit that involves a linked Visa card can allow attackers to steal money from a locked device using NFC, but the process is complex, requiring physical access and specialized hardware. The exploit was highlighted by popular YouTube channel Veritasium, and it involves tricking an iPhone into thinking it's making a payment at a mass transit terminal, a process that can be completed from a locked iPhone.


Cybersecurity researchers from the University of Surrey and the University of Birmingham developed the attack to bypass an iPhone's locked status and steal funds from a mobile wallet. The exploit was first publicized in 2021, and it bypasses traditional limits on transaction size. Veritasium demonstrated the attack by collecting $10,000 from YouTuber Marques Brownlee's locked iPhone.

The attack works using an NFC card reader that intercepts the communication between an iPhone and a tap-to-pay terminal when a payment is made. The card reader is connected to a laptop that collects payment data and sends it to a separate burner phone, which is then tapped on a legitimate card reader. The NFC device has to be tuned to the same transit terminal identifier as a legitimate transit reader.

The process requires the victim to have Express Transit Mode enabled for payments, and a Visa card linked for those payments, among other steps. As it turns out, it's a Visa-related security loophole rather than an iPhone issue, and it doesn't work with a Mastercard or an American Express card because other cards use different security methods. It also doesn't work with Samsung Pay on Samsung devices, and it requires the specific combination of a Visa card and an iPhone. Apple told Veritasium that it's an issue with the Visa system, but something unlikely to occur in the real world.
Visa also told Veritasium that the exploit was very unlikely from a scaled real world setting, and any such transactions can be disputed. The researchers who shared the exploit said users can protect themselves by not using a Visa card on the iPhone for transit purposes.
This article, "Here's How Researchers Stole $10,000 From MKBHD's Locked iPhone" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Security vendor Pluto Security has published details of a critical vulnerability in the open-source nginx UI web server configuration tool that has been under active exploitation by cybercriminals since March.
News of the flaw, identified as CVE-2026-33032, first appeared on the National Vulnerability Database (NVD) on March 30, the same day that threat intelligence companies VulnCheck and Recorded Future’s Insikt Group noted it was under active exploitation.
What users didn’t have at that point were any details on the flaw from Pluto Security, the company that discovered it earlier that month. This week, the company rectified this, publishing a full breakdown of the vulnerability.
Nginx UI is a convenient real-time dashboard and control panel interface for managing nginx single-node and cluster nodes without having to resort to the command line interface (CLI).
The vulnerability, with a CVSS score of 9.8, relates to the software’s support for Model Context Protocol (MCP) servers, which was added in late 2025 and enables communication between nginx web servers and AI models though two HTTP-accessible MCP URL endpoints.
Unfortunately, in the case of nginx UI, one of these endpoints, /mcp_message, was implemented without authentication, a weakness Pluto Security dubbed ‘MCPwn’.
“This exposes 12 MCP tools, including config writes with automatic nginx reload, to any host on the network. One unauthenticated API call is all it takes to inject a config and take over nginx,” said Pluto Security.
Leveraging MCPwn, an attacker would be able to intercept all traffic, harvest admin credentials, maintain persistent access, conduct infrastructure reconnaissance via nginx configuration files, and kill the service, the company said.
MCP attack surface
Nginx UI’s user base of hundreds of thousands is relatively small compared to the vast global popularity of the nginx web server. Many of its installations will also be internal and therefore not directly exposed to remote attack. However, using Shodan, Pluto Security was still able to find 2,689 vulnerable nginx UI instances reachable from the internet, it said.
“This is a clear example of how AI integrations can unintentionally expand the attack surface,” commented Pluto Security’s CEO, Shahar Bahat. “MCP servers aren’t just developer tools, they’re privileged access points into production systems.”
MCP has been implemented at breakneck speed to enable AI agents, leading to the adoption of tools without the risks they create being understood, Bahat pointed out.
“This vulnerability shows how a single exposed endpoint can enable full compromise. AI integration layers must be treated as part of the attack surface, not an afterthought,” she said.
To security teams, this will be reminiscent of the problems experienced when APIs started to boom a decade ago. By enabling an integration layer such as MCP, and the tools used to manage it, developers risk inadvertently creating a new layer of vulnerability.
As Bahat put it: “AI integration endpoints expose the same capabilities as the core application, but often skip its security controls.” When planning MCP integrations, Pluto Security recommends giving MCP endpoints the same security attention as APIs, auditing Server-Sent Events (SSE) endpoints and fully testing authentication parameters.
A priority fix
The fact that the nginx vulnerability has been under exploitation for at least a month should make applying the recommended fix, version 2.3.4, released March 15, a priority for anyone using this software, since nginx servers represent a big prize for threat actors. In February, attackers were discovered exploiting the ‘React2Shell’ vulnerability (CVE-2025-55182) inReact Server Components (RSC) to target nginx servers.
For those who can’t patch immediately, the stopgap workaround is to disable MCP, or lock the IP whitelist to trusted hosts, as well as reviewing access logs for unusual configuration changes.

View the full article
Apple is sending a large portion of its Siri engineers to a multi-week bootcamp to learn to code using AI, reports The Information. Apple's decision to teach its programmers to better use AI for coding comes just two months before Apple is expected to unveil a smarter, more capable version of Siri at WWDC.


While employees attend the coding bootcamp, around 60 members of the ‌Siri‌ development team will stick around to work on ‌Siri‌, and an additional 60 will evaluate how ‌Siri‌ is performing. Apple is testing to make sure ‌Siri‌ is meeting its safety standards and is able to interpret and execute commands from users.

Coding with AI is becoming the standard, but Apple's ‌Siri‌ team apparently isn't taking full advantage of AI coding tools. The Information says that some teams within Apple have allocated large parts of their budgets to Claude Code, but the ‌Siri‌ team has a "reputation as a laggard inside Apple."

The ‌Siri‌ team was unable to produce the Apple Intelligence version of ‌Siri‌ that Apple promised would come in iOS 18, leading to a major organizational shakeup. Apple replaced AI chief John Giannandrea, who stepped down from his position in late 2025 and is set to retire this week following the final vesting of his stock on April 15.

Apple software engineering chief Craig Federighi took over and oversees AI development, and Mike Rockwell, who developed the Vision Pro, is the Siri team lead. Under Federighi, Apple inked a deal with Google that will see ‌Siri‌ and other AI features powered by Google's Gemini models.Tag: Siri
This article, "Siri Engineers Sent to AI Coding Bootcamp as Apple Prepares to Deliver Siri Overhaul" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Google is bringing Gemini to the Mac with a new native macOS app that's available starting today. Gemini for Mac can be activated with a keyboard shortcut, and it has built-in tools for generating images, analyzing what's on your screen, reviewing files, and more.


Gemini is the last of the three major AI services to have a dedicated Mac app, because OpenAI and Anthropic have had Mac apps for quite some time.

Gemini can be brought up anywhere on the Mac with an Option + Space keyboard command, so there's no need to swap over to a dedicated window. Option + Shift + Space is available for opening the full Gemini chat window. Gemini can also be accessed from the Dock or through the Menu Bar.

Any window on the Mac can be shared with Gemini, allowing Gemini to provide contextual assistance on anything that you're looking at. After activating Gemini, selecting the Share Window option will let Gemini see what it is you want to ask questions about. Gemini will need Accessibility access to read full pages in a browser window.

Nano Banana is available for creating images, and Veo can be used for generating videos.

Gemini for Mac is available for Macs running macOS 15 and later, and it is free to download and use. Free access to Gemini is limited, and Google has subscription plans with increased usage limits. Google AI Plus is $7.99 per month, Google AI Pro is $19.99 per month, and Google AI Ultra is $249.99 per month.

Google says that the Mac app is the first step toward a personal, proactive, and powerful desktop assistant, with more news to follow in the coming months.Tags: Gemini, Google
This article, "Google Launches Native Gemini AI App for Mac" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Threat actors have been observed weaponizing n8n, a popular artificial intelligence (AI) workflow automation platform, to facilitate sophisticated phishing campaigns and deliver malicious payloads or fingerprint devices by sending automated emails. "By leveraging trusted infrastructure, these attackers bypass traditional security filters, turning productivity tools into deliveryView the full article
According to the latest rumors, Apple is close to launching its next-generation iPad mini. So what should we expect from the successor to the iPad mini 7 that Apple released over a year ago? Read on to find out.


Processor and Performance

Apple is working on a next-generation version of the iPad mini (codename J510/J511) that features the A19 Pro chip, according to information found in code that Apple mistakenly shared in August.

Apple's A19 Pro chip since debuted in the iPhone Air and iPhone 17 Pro models. The iPhone 17 Pro models include the higher-end version of Apple's A19 Pro chip with a 6-core CPU and a 6-core GPU, while the iPhone Air uses a mid-tier A19 Pro chip with one fewer GPU core than the A19 Pro chip used in the iPhone 17 Pro and Pro Max.

If the code leak is accurate for the iPad mini 8, Apple is likely to use the mid-tier A19 Pro chip found in the iPhone Air. This is based on the fact that the A17 Pro chip used in the iPad mini 7 has a 6-core CPU with two high-performance cores and four efficiency cores, along with a 5-core GPU, compared to the 6-core GPU found on the A17 Pro used in the iPhone 15 Pro.

Apple built the A19 Pro chip on an upgraded third-generation 3-nanometer N3P process for modest speed and efficiency improvements. The chip includes a 16-core Neural Engine, next-generation dynamic caching, and unified image compression.

The GPU in the A19 Pro has an upgraded architecture with a larger cache, more memory, and Neural Accelerators that are built into each core. Apple says that this change provides 3× the peak GPU compute over the prior-generation chip. There's also an upgraded 16-core Neural Engine for AI tasks.

There is an outside chance that Apple opts for the A20 Pro chip for the new iPad mini. The claim has been made by a MacRumors tipster who analyzed a macOS kernel debug kit containing internal Apple codenames. However, the iPad mini has not always received Apple's newest A-series chip at the time it was updated, so the A19 Pro cannot be ruled out at this time. iPhone 18 Pro models are also expected to use the A20 Pro chip, which will reportedly be fabricated with TSMC's advanced 2nm process.

Display


Apple's plan to transition the ‌‌iPad mini‌‌ from an LCD to an OLED display is widely rumored. According to Bloomberg's Mark Gurman, the small form-factor tablet is likely to be the next Apple device to adopt OLED. According to a Chinese leaker with sources in Apple's supply chain, Apple has evaluated a Samsung-made OLED display for its next iPad mini model.

It remains unclear whether the iPad mini 8 will feature a higher refresh rate than the 60Hz LCD display used in the existing iPad mini 7, but since the new base iPhone 17 now uses a 120Hz ProMotion panel, it would be reasonable to expect the same on the first OLED iPad mini. A separate report has suggested the ‌‌‌iPad mini 8‌‌‌'s screen could increase in size from 8.3 inches to 8.7 inches with the adoption of OLED.

OLED panels can individually control each pixel, resulting in more precise color reproduction and deeper blacks compared to other common display technologies. They also provide superior contrast, faster response times, better viewing angles, and greater design flexibility. All of Apple's flagship iPhones use OLED panels, and in May 2024 the company brought the display technology to the iPad Pro for the first time.

Unlike Apple's ‌iPad Pro‌ models, which feature two-stack low-temperature polycrystalline oxide (LTPO) OLED panels‌, the ‌iPad mini‌ may have a single-stack low-temperature polycrystalline silicon (LTPS) panel, which would make it dimmer.

Chassis Design


Apple is reportedly working to give the iPad mini 8 a more water-resistant design, according to Bloomberg's Mark Gurman. The updated casing would bring protection levels closer to those of the iPhone, making the tablet safer for use in damp environments.

To achieve this, Apple is said to have designed a new vibration-based speaker system that eliminates the need for traditional speaker holes. By using sound-emitting surfaces instead of open grilles, the company can reduce potential entry points for water and dust, resulting in a more sealed, durable enclosure.

On the iPhone, Apple relies on adhesives and gaskets to shield speakers and other openings from moisture. The iPad mini's approach appears to go further, doing away with the holes altogether. Current iPad mini models lack any official IP rating, but the upcoming version could mark the first in the lineup to feature a certified level of water protection.

Apple patents could offer further clues to the new design direction. For example, a 2014 patent outlines a "mechanically actuated panel acoustic system" that vibrates flat surfaces to generate sound, effectively turning parts of a device's chassis into a speaker diaphragm. This could potentially allow Apple to produce audio without visible speaker holes. The patent suggest Apple has been building towards a sealed, vibration-based acoustic system for several years.

Release Date


According to research firm Omdia, the ‌‌iPad mini‌‌ is expected to adopt an OLED display in 2027. However, Korea's ET News and ZDNET Korea have both suggested that the iPad mini will be updated with an OLED display in 2026. Bloomberg has also said the update could come as soon as this year.

The most recent word on the subject comes from Weibo-based leaker Instant Digital, who claims the OLED iPad mini will be launched in the second half of 2026 at the earliest.

In May 2024, it was reported that Samsung Display had started developing sample OLED panels for a future ‌iPad mini‌, with plans to initiate mass production at its facility in Cheonan in the second half of 2025. The same report claimed that Apple will bring an OLED panel to the iPad Air alongside the ‌iPad mini‌ in 2026, though Apple only refreshed the iPad Air in March, and more recent reporting suggests an OLED iPad Air will arrive in early 2027.

The latter outlook aligns with a December report by analyst firm Display Supply Chain Consultants (DSCC) that said an 8.5-inch OLED iPad mini is planned for a 2026 launch, while 11-inch and 13-inch OLED iPad Air models are expected to follow in 2027.

Ultimately, there are no rumors suggesting exactly when the next ‌iPad mini‌ will be released, but a launch later in 2026 has a high probability.

Pricing


Apple's ‌iPad mini‌ with OLED display technology and improved water resistance is expected to be more expensive, and Apple could charge up to $100 more for the device, according to Bloomberg's Gurman. The ‌iPad mini‌ is currently priced starting at $499. Gurman has previously argued that Apple should consider a lower-end version of the mini, or at least a change to its current $499 starting price, given that it's up against rival products that cost a lot less.

However, Apple users who are looking for a more affordable option should probably consider the 10th-generation iPad instead. Starting at $329, the iPad offers many iPad mini features, such as Touch ID and Center Stage, but at a lower price that balances functionality and affordability. Related Roundup: iPad miniTag: OLEDBuyer's Guide: iPad Mini (Caution)Related Forum: iPad
This article, "OLED iPad Mini: Release Date, Pricing, and What to Expect" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Apple today announced a series of events tied to this year's TCS London Marathon, with the company serving as an official partner of the race.


The TCS London Marathon is one of the world's most popular marathons and takes place on Sunday, April 26, drawing athletes of all abilities from around the world. A record-breaking one million people applied to enter the ballot for the 2026 event.

Apple will host two free events at its Brompton Road store in the days before the race. On Thursday, April 23, a panel including fitness trainer Joe Wicks, ultramarathon runner Hellah Sidibe, and athletes Dora Atim, Becky Briggs, and Sherica Holmon will offer training tips before a 5K shakeout run in Hyde Park, hosted by Apple Fitness+ trainer Cory Wharton-Malcolm. Spots are limited and registration is now open.

On Saturday, April 25, former marathon world-record holder Paula Radcliffe and two-time Olympian Chris Thompson will record a live episode of their podcast, Paula's Run Club, also at Brompton Road, joined by Wharton-Malcolm. The episode closes out their "Road to London Marathon" series. Registration is available for that event as well.

On race day, Apple Music will station artists and DJs at a key point on the course. An official Ultimate Marathon Playlist spanning seven hours is available now on ‌Apple Music‌, with additional mixes from race-day DJs to follow after the event.

Ahead of marathon week, Apple will host a PE with Apple: Hour of Play event for students from six schools in the London borough of Wandsworth, with Wicks and Fitness+ trainers leading physical activities for children ages 10 to 14, in partnership with nonprofit Enable. Apple also pointed out that it supports several other Greater London organizations, including Battersea Arts Centre, Southbank Centre, Youth Battersea, and Wandsworth BEST.Tag: United Kingdom
This article, "Apple to Host Free Events in London Ahead of Sunday's Marathon" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Amazon recently introduced fresh deals on the Apple Watch Ultra 3, providing $99 discounts on select models. These are the best prices on the Ultra 3 that we've tracked so far in 2026, and they're overall solid second-best prices.

Note: MacRumors is an affiliate partner with some of these vendors. When you click a link and make a purchase, we may receive a small payment, which helps us keep the site running.

You can get the Apple Watch Ultra 3 for $699.99 in the Natural color option, down from $799.00. There are also a few Milanese Loop models on sale for $799.99, down from $899.00.

$99 OFFApple Watch Ultra 3 for $699.99

We've collected all of the Apple Watch Ultra 3 models currently on sale on Amazon in the list below. All of these deals are within $19 of the all-time low price, and they're the best prices we've seen so far in 2026.

Natural with Anchor Blue Ocean Band - $699.99 ($99 off)
Natural with Blue/Bright Blue Trail Loop (M/L) - $699.99 ($99 off)
Black with Black Milanese Loop (Large) - $799.99 ($99 off)
Natural Milanese Loop (Large) - $799.99 ($99 off)

If you're on the hunt for more discounts, be sure to visit our Apple Deals roundup where we recap the best Apple-related bargains of the past week.



Deals Newsletter

Interested in hearing more about the best deals you can find in 2026? Sign up for our Deals Newsletter and we'll keep you updated so you don't miss the biggest deals of the season!




Related Roundup: Apple Deals
This article, "Apple Watch Ultra 3 Gets Lowest Prices of 2026 So Far With $99 Off Select Models" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Apple privately warned Elon Musk's xAI company in January that it would remove the Grok app from the App Store unless the company put a stop to the chatbot's nude and sexualized deepfakes, according to a letter Apple sent to U.S. senators and obtained by NBC News ($).


Earlier this year, Grok's AI capabilities came under scrutiny after X users shared nonconsensual sexualized images of women and children created by the app, many of which were based on photos of real people.

What followed was a confusing rollout of moderation changes to Grok, some of which could be easily bypassed. Publicly, Apple did not comment on the controversy at the time, but it did respond, and was in fact the instigator of the changes. Internally, the company had found both X and Grok in violation of its App Store guidelines and demanded its developers submit a content moderation plan, the letter reveals.

According to the letter, Apple rejected an initial fix from xAI as insufficient, saying the "changes didn't go far enough," and Apple warned it that additional alterations were required or Grok would be removed. After further back-and-forth, however, Apple eventually concluded that a later submission of the app had improved enough for it to be approved.

The disclosure was apparently prompted by a January letter from Senators Ron Wyden, Ben Ray Luján, and Edward Markey, who urged Apple and Google to pull both apps, arguing the imagery violated App Store rules barring offensive, sexual, and exploitative content.

The senators also said that Apple's response would test its own arguments, since the company has long defended its curated App Store by claiming its review process keeps users safer. Letting Grok continue to generate this kind of imagery, they argued, would undermine that case in the eyes of the public and in a court of law.

After NBC News published its report, X posted the following statement on its platform:
While the amount of sexualized deepfakes created by Grok and posted to X appears to have decreased significantly, NBC News found that Grok is still able to generate similar imagery, with some users apparently having simply updated their prompt tactics to get around the safeguards. You can read that report in its entirety by following this link.Tags: App Store, Grok
This article, "Apple Threatened to Pull Grok From App Store Over Sexualized Images" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Enterprise AI agents are supposed to streamline workflows. Instead, two fresh findings show they can just as easily streamline data exfiltration.
Security researchers have uncovered prompt-injection vulnerabilities in both Microsoft Copilot Studio and Salesforce Agentforce that allow attackers to execute malicious instructions via seemingly harmless prompts.
According to Capsule Security findings, SharePoint forms and public-facing lead forms within Copilot are vulnerable to attackers issuing prompts that can override system intent and trigger data exfiltration to attacker-controlled servers.
One of these flaws has already been assigned a high-severity CVE, with another “critical” one reportedly missing the bar for categorization. The flaws can allow theft of PIIs, customer/lead records, free-text business context, and operational/workflow data.
In both cases, AI agents treat untrusted user input as trusted instructions, Capsule researchers noted in the disclosures shared with CSO ahead of their publication on Wednesday.
ShareLeak: SharePoint forms data leaked through Copilot
The Microsoft-side issue, dubbed “ShareLeak,” is about how Copilot Studio agents process SharePoint form submissions. The attack begins with a crafted payload inserted into a standard form field, like “comments”, which the agent later ingests as part of its operational context.
Because the system concatenates user input with system prompts, the injected payload overrides the agent’s original instructions. The model is thus tricked into believing the attacker’s instructions are legitimate system directives. The malicious input moves from form submission to agent execution without any resistance.
Once compromised, the agent can access connected SharePoint Lists and extract sensitive customer data, including names, addresses, phone numbers, and send it externally via email. The researchers found that even when Microsoft’s safety mechanisms flagged suspicious behavior, the data was exfiltrated.
The root cause is that there is no reliable separation between trusted system instructions and untrusted user data. In the existing setup, the AI cannot distinguish between the two, the researchers said.
Microsoft patched the issue following disclosure, assigning CVE-2026-21520 to it and assessing its severity at 7.5 out of 10 on the CVSS scale. The mitigation was carried out internally, and no further action is required from the users.
PipeLeak: Salesforce Agentforce hijacked by a simple lead
In the Salesforce Agentforce case, attackers embed malicious instructions inside a public-facing lead form. When an internal user later asks the agent to review or process that lead, the agent executes the embedded instructions as if they were part of its task.
According to a Capsule demonstration, the agent retrieves CRM data via the “GetLeadsInformation” function and then sends it externally via email.
The compromise isn’t limited to a single record. Researchers demonstrated that a hijacked agent could query and exfiltrate multiple lead records in bulk, effectively turning a single form submission into a database extraction pipeline.
The researchers said Salesforce acknowledged the prompt injection issue but characterized the exfiltration vector as “configuration-specific,” pointing to optional human-in-the-loop (HITL) controls. Capsule’s pushback on that framing argues that requiring manual approvals undermines the very purpose of autonomous agents.
The deeper issue, they noted, is insecure defaults. Systems designed for automation should not allow untrusted inputs to redefine agent goals.

Both disclosures converge on a baseline that calls for treating all external inputs as untrusted and having filters in place that separate data from instructions. This would entail enforcing input validation, least-privilege access, and strict controls on actions like outbound email.
View the full article
Apple will bring OLED displays to its iPad Air models next year, according to a new report from Korea's ET News.


Citing industry sources, the outlet says Samsung Display will begin mass production of OLED panels around the end of 2026 or January next year, with a view to supplying panels for Apple's next iPad Air, expected to be released in early 2027. Apple last updated the iPad Air in March 2026 with an M4 chip.

Apple's iPad Pro models already have OLED displays, but the iPad Air models still use more affordable LCD displays that Apple calls Liquid Retina. The Liquid Retina displays do not support 120Hz ProMotion display technology, and are limited to 60Hz refresh rates.

OLED panels individually control each pixel, resulting in more precise color reproduction and deeper blacks compared to LCD. They also provide superior contrast, faster response times, better viewing angles, and greater design flexibility.

That said, unlike Apple's ‌iPad Pro‌ models, which feature two-stack low-temperature polycrystalline oxide (LTPO) OLED panels‌, the iPad Air‌ is expected to use single-stack low-temperature polycrystalline silicon (LTPS) panels, meaning that they may be dimmer and continue to lack ProMotion.

Apple's plan to transition the ‌‌iPad mini‌‌ from an LCD to an OLED display is already widely rumored, with reports suggesting the iPad mini 8 will adopt OLED later this year, albeit using the same cheaper single-stack LTPS panel.

Once the iPad mini and iPad Air receive the display upgrade, the entry-level iPad will be the only model in Apple's tablet lineup without an OLED panel.Related Roundup: iPad Air Tags: ETNews, OLEDBuyer's Guide: iPad Air (Buy Now)Related Forum: iPad
This article, "Report: iPad Air to Gain OLED Display Next Year" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Few technologies have moved from experimentation to boardroom mandate as quickly as AI. Across industries, leadership teams have embraced its broader potential, and boards, investors, and executives are already pushing organizations to adopt it across operational and security functions. Pentera’s AI Security and Exposure Report 2026 reflects that momentum: every CISO surveyedView the full article
Anthropic has released a redesigned Claude Code experience for its Claude desktop app, bringing in a new sidebar for managing multiple sessions, a drag-and-drop layout for arranging the workspace, and more.


The new sidebar displays every active and recent session in one place, and users can filter by status, project, or environment, with the option to group sessions by project. A new side chat shortcut (Command + ;) also lets users branch a question off a running task without feeding extra context back into the main thread.

Anthropic has also dropped more of the developer workflow into the app itself. There's now an integrated terminal for running tests and builds, an in-app file editor for spot edits, a rebuilt diff viewer aimed at large changesets, and an expanded preview pane that handles HTML files and PDFs alongside local app servers. Each pane is also drag-and-drop friendly, so the layout can be arranged to suit.

In addition, the desktop app now matches Claude Code's CLI for plugin support, while SSH sessions are supported on Mac as well as Linux. Lastly, there are three view modes (Verbose, Normal, and Summary) that let users decide how much of Claude's tool-call activity they want to see.

The Claude desktop app update is rolling out now to Claude Code users on Pro, Max, Team, and Enterprise plans.


In related news, Anthropic also announced Routines – a new way to set up Claude Code automations that run without an active session. A routine bundles a prompt, a repo, and any relevant connectors into a single configuration that can run on a schedule, fire from an API call, or trigger off a GitHub event such as a new pull request.

Routines run on Claude Code's web infrastructure rather than a local machine, but Anthropic has put in place daily run caps that scale by plan. Routines are available in research preview to Pro, Max, Team, and Enterprise users with Claude Code on the web enabled. For further details, check out Anthropic's blog post.Tag: Anthropic
This article, "Anthropic Rebuilds Claude Code Desktop App Around Parallel Sessions" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Deepfake technology has crossed a critical threshold. What was impossible 10 years ago and required specific expertise only a few years ago is now cheap and accessible. Worse, it’s now good enough to fool a wide range of employees and executives. In fact, a 2025 Gartner survey found that 43% of cybersecurity leaders experienced at least one audio deepfake, and 37% experienced video deepfakes in the past year.
These findings reflect what we see in the wild. Deepfakes are no longer a hypothetical future risk; they are showing up in real workflows, decisions and incidents. Meeting this sophisticated and evolving challenge demands rapid verification, cross-functional coordination and clear communication to limit the impact of synthetic media.
The collapse of heuristics
Humans tend to rely on mental shortcuts, or heuristics, to judge what is real. Historically, trusting a familiar face or voice was enough to feel confident. Early deepfakes began to test these instincts, but they were often marred by telltale flaws, such as unusual blinking, mangled fingers, or blurred text.
Unfortunately, that tried-and-true approach to distinguishing fact from fiction is no longer reliable. The obvious cues that made deepfakes easier to spot have been eliminated by the latest generation of generative AI models, such as Nano Banana Pro. As a result, the heuristics humans have historically relied on cannot be trusted in isolation.
Deepfakes as tools for financial fraud
Deepfakes have quickly become a powerful enabler of financial fraud. This is largely because most business communication channels, like video and voice calls, remain unauthenticated. A single convincing audio or video call, seemingly from a trusted executive, can bypass established controls in minutes. Employees in these scenarios often follow instructions or approve large fund transfers, believing they are acting on legitimate requests.
A well-known example of this risk occurred at Arup, where the company’s CFO and other video call participants were convincingly simulated using AI-generated deepfakes, and an employee transferred roughly $25 million.
Until robust dual authentication for phone calls is standard, organizations remain exposed to anyone who can convincingly mimic a CFO or CEO.
Deepfakes as reputational weapons
Financial fraud remains a serious concern with deepfakes, but they are increasingly being used as reputational weapons, engineered to erode confidence among investors, customers and business partners. Attackers need only a brief clip, often as little as 20 seconds, to impersonate an executive and unravel years’ worth of reputation and trust built with key stakeholders. Beyond the C-cuite, anyone with a digital footprint, from a podcast appearance to a short social media clip, could become a target.
Recent cases show how rapidly these false narratives can escalate and cause real damage:
Market destabilization: In January 2026, the Bombay Stock Exchange was forced to issue an urgent warning after deepfake videos of its CEO spread online, promoting fraudulent stock tips and promises of “supernormal profits.” Public disruption: After a December 2025 earthquake in the UK, a synthetic image of a collapsed bridge went viral, leading to train cancellations. Internal sabotage: In a private case, a former employee created deepfakes of company leaders making inflammatory remarks and distributed them directly to business partners, intending to inflict reputational damage. Each of these incidents forced the affected organizations into crisis mode. The rapid spread of deepfakes on digital platforms means false content often circulates faster than teams can investigate or respond to it. By the time the truth emerges, the damage to relationships and reputation may already be done.
Building resilience against deepfakes
Deepfake incidents differ from other cyber attacks. While they may not cause immediate financial loss, they often unfold publicly, spread faster than investigations can keep pace and exploit human trust at scale. For most organizations, handling the widespread uncertainty and reputational damage stemming from a deepfake incident exceeds the capabilities of internal teams, especially when public trust is at stake.
Addressing this challenge requires more than technical controls. Business leaders are increasingly recognizing the importance of being able to respond to these threats quickly and decisively. Effective response now depends on capabilities that enable organizations to verify content, limit its spread and communicate with stakeholders in a timely and credible way. In practice, this includes:
Technical analysis: Expert forensic review of audio and video content to determine whether the content has been manipulated and to generate forensic proof for stakeholders. Legal support: The ability to act once harmful content has been identified, including coordinating takedown requests by working with legal experts to support the removal of malicious or defamatory content from online platforms. Clear communication: Public relations and communications support to help organizations craft effective messages for employees, investors and customers during a rapidly evolving incident. The path forward: Authentication as the end state
In the long term, addressing deepfakes will likely require broad adoption of authentication and watermarking standards, like how web browsers display a lock icon to signal a secure, authenticated connection. For example, organizations may soon embed watermarks in official communications, such as press statements, interviews and earnings calls.
Yet, watermarks will not resolve every challenge. Some authentic content, like revelations from whistleblowers, will inevitably circulate without official marks. Attackers will still be able to fake this kind of content, leaving us in a continual cat-and-mouse game, in which journalists and forensic experts must draw on alternative sources and advanced tools to verify materials. Establishing trust in digital media will remain an ongoing process, as both attackers and defenders adapt.
For business and risk professionals, the takeaway is clear: True resilience no longer depends on heuristics and trusting what we see or hear. It depends on how quickly organizations can verify reality, coordinate a response with expert support and resources and restore trust before misinformation becomes the dominant narrative.
This article is published as part of the Foundry Expert Contributor Network.
Want to join?
View the full article
Apple and Amazon are partnering up for a $19.99/month streaming TV bundle that includes access to Apple TV and Peacock Premium Plus.


Available to customers in the U.S. for a limited time, the streaming bundle offers access to both services at a 30% discount, similar to the bundle that launched in October 2025.

‌Apple TV‌ is normally $12.99 per month, while Peacock starts at $16.99 for the Premium plan that's included in the bundle, so that's a $10 monthly saving. Peacock Premium Plus includes the option to download content and watch it offline. It also removes most ads.

The Apple TV and Peacock Premium Plus bundle offers customers access to live sports like Premier League soccer and Major League Baseball, Apple Originals like Pluribus and Severance, as well as Peacock shows and moves. Customers can watch content from both services directly through the Prime Video app on compatible devices.
Prime members can subscribe to the bundle via the Prime Video app or visiting the Prime Video website and using their existing Amazon account and payment method.Related Roundup: Apple TVBuyer's Guide: Apple TV (Don't Buy)Related Forum: Apple TV and Home Theater
This article, "Apple TV and Peacock $20 Monthly Bundle Available on Amazon Prime Video, Offering a 30% Discount" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Cyberattacks targeting the healthcare sector have surged since the COVID-19 pandemic and the resulting rush to enable remote delivery of healthcare services. Security vendors and researchers tracking the industry have reported a major increase in phishing attacks, ransomware, web application attacks, and other threats targeting healthcare providers.
Recent rising of ransomware attacks on healthcare, in particular the Change Healthcare breach, has been a headline-grabbing wake-up call for healthcare execs.
The trend has put enormous strain on healthcare security organizations. “The healthcare industry is under siege from a range of complex security risks,” says Terry Ray, vice president of product strategy at Varonis. “Cybercriminals are hunting for the sensitive and valuable data that healthcare has access to, both patient data and corporate data.”
Many organizations are struggling to meet the challenge because they are under-resourced and rely on vulnerable systems, third-party applications, and APIs to deliver services.
Moreover, IT systems are increasingly used to optimize clinical encounters and patient care. Implantable devices, such as loop recorders, are increasingly being used to aid diagnoses, for example, of cardiac arrhythmias. These devices support telemetry, as do wearable devices, by transmitting patient data. Important healthcare decisions are made based on this data, with patient data being far more available due to advances in IT.
The increasing usage of IoT and IT in healthcare have improved clinical efficiency and decision-making certainty, but it does require greater attention to risk assessment, with the days of patient data stored in locked filing cabinets long gone, says WithSecure principal consultant Stuart Morgan.
“The impact of patient data being manipulated or leaked is intuitive and well understood, but the risk of denial of service — whether malicious or unintended — can be huge,” Morgan tells CSO. “Although resilience is built in to these systems to a degree, resorting to backup systems are by their nature far less efficient.”
Here, security experts identify the major cybersecurity threats healthcare organizations face today.
The rising ransomware threat
Ransomware has emerged as one of the biggest cyber threats for healthcare today. Attackers have discovered that healthcare organizations delivering life-saving treatments can be more easily extorted than victims in almost every other sector. Many healthcare organizations are also more susceptible to attacks because of new digital applications and services they have launched to address demand for telehealth services, among other digitalization efforts.
From 2022 to 2023, healthcare ransomware victims jumped 81%, according to a study by US Office of the Director of National Intelligence. This past year, healthcare ransomware attacks increased another 30%, as vendors and service partners joined clinics and hospitals as key targets for attacks, according to a study by Comparitech.
Pharmaceutical manufacturers, medical billing providers, and healthcare tech companies have also come under increasing fire from ransomware actors, Comparitech found.
Change Healthcare’s devastating ransomware attack in February 2024 is among the most notable. The attack, which disrupted insurance claim processing, prescription dispensing, and financial settlements, had a huge impact on hospitals, clinics, and pharmacies across the US. In August 2024, Michigan-based McLaren Health Care suffered the second of two ransomware attacks over the course of just 12 months.
A June 2024 ransomware attack on NHS-affiliated UK pathology services Synnovis caused massive disruption in parts of London, forcing hospitals to cancel planned procedures and resulting in a temporary shortage of blood supplies. The Qilin ransomware group exfiltrated data before deploying ransomware that hobbled Synnovis’ IT systems, affecting blood tests, diagnostics, and lab services.
Medical equipment firm Stryker experienced global network disruptions to its IT systems following a cyberattack in March 2026. Initially ransomware was suspected but Iran-linked group Handala quickly became the prime suspect in an attack that wiped an estimated 200,000 devices as part of a broader campaign against US and Israeli targets.
Electronic health records (EHRs) and systems present the biggest risk in healthcare today, says Caleb Barlow, president and CEO of CynergisTek. “Past attacks have shown when a hospital undergoes a ransomware-induced lockdown period, access to EHRs is shut down, and patients may have to be diverted for care,” he says. “Such attacks can prevent access to critical prescription information and dosing for patients with complex, chronic conditions like diabetes or cancer. Worse, hackers can potentially take it a step further and manipulate health record data to undermine patient care.”
Historically, healthcare institutions transferred this risk to cyber insurance, but that is becoming more difficult because insurers are making it harder for organizations to purchase ransomware protection without specific controls such as multi-factor authentication and endpoint detection and response technologies, Barlow says.
Cloud vulnerabilities and misconfigurations
Many healthcare organizations have adopted cloud services as part of broader digital transformation initiatives. As a result, patient health information (PHI) and other sensitive data is increasingly being hosted in vendor cloud environments.
The trend has broadened attack surface at healthcare organizations, says Anthony James, vice president of products at Infoblox, especially as healthcare organizations often use multiple cloud vendors and services with differing security standards and practices, making it hard to apply a consistent data protection policies.
Sixty-one percent of healthcare companies said they experienced a cloud cyberattack in the past 12 months in a February 2024 report by healthcare software developer KMS Healthcare.
In March 2026, US healthcare software vendor CareCloud’s EHR environment suffered a breach, disrupting access for 45,000 providers.
Attacks aren’t the only cyber risks healthcare organizations face with rising cloud use. Misconfigurations play a role as well.
In April 2025, US health insurer Blue Shield of California found that it had exposed member data — including protected health information — to Google’s advertising platform for three years up until January 2024 because of a flawed Google Analytics setup on some of its web pages.
Web application attacks
Web application attacks targeting healthcare entities have also spiked sharply in recent years, with cross-site scripting attacks among the most common, along with SQL injection, protocol manipulation attacks, and remote code execution/remote file inclusion attacks.
“Technically speaking, web application attacks can be incredibly challenging for under-resourced healthcare organizations to manage,” Varonis’ Ray says. To address the issue, healthcare organizations must implement controls that enable better visibility into third-party applications and API connections. Only then will the security team be able to understand who is trying to access critical data and whether that activity should be permitted.
Bad-bot traffic
Traffic from bad bots — such as those attempting to scrape data, send spam, or download unwanted software — present another major challenge for healthcare organizations. The problem became especially pressing during the pandemic when governments around the world set up new websites and other digital infrastructure to support COVID vaccine registrations and appointments.
“Increased levels of traffic result in downtime and disruption for legitimate human users who are trying to access critical services on their healthcare providers’ site,” Ray says. “It might also result in increased infrastructure costs for the organization as it tries to sustain uptime from the persistent, burdensome level of elevated traffic.”
The latest 2025 edition of Imperva’s Bad Bot report estimates malign bots account for nearly a third (37%) of internet traffic, up from 32% in the year prior. Imperva warned that AI is “supercharging the bot threat” alongside a shift in advanced bot traffic targeted APIs rather than applications, reflecting how API endpoints often handle sensitive or high-value data.
“Financial services, business, telecom, and healthcare are among the most targeted industries for bot attacks on APIs, accounting for over 75% of all API attacks,” Imperva reports.
Bad bots can lead to healthcare data breaches, for example through credential stuffing attacks against patient accounts, and scraping of sensitive health information.
Cybercriminals target confidential health information, such as patient records, medical history, and insurance details because this stolen data can be sold on the dark web for profit or used for fraudulent activities, Imperva warns.
Increased phishing volumes
Phishing attacks pose a major threat to the healthcare industry as they do in almost every sector. Again, the pandemic provided a unique backdrop for a rise in phishing volumes versus healthcare organizations. In a survey of 168 healthcare cybersecurity professionals conducted by Healthcare Information and Management Systems Society (HIMSS) at the time found that phishing was the typical initial point of compromise for most security incidents.
“Phishing attacks are the top type of significant security incident reported by respondents,” HIMSS noted in its report. “Phishers were the top type of threat actor responsible for significant security incidents at healthcare organizations.”
But phishing has long been an issue for healthcare. Stats compiled by the US Department of Health and Human Services (HHS) record that 18% of 4,419 reported breaches of PHI between 2009 and 2021 involved either phishing attacks or the hacking of email accounts, according to the HIPAA Journal.
Phishing was the initial vector in high-profile attacks against healthcare organizations Anthem (2015) and Magellan Health (2020), among others.
A study by UK medical journal BMJ found that around 3% of emails sent to hospital staff over a one-month period were suspected threats.
While many staff appear to be aware of phishing and respond appropriately, ongoing education is required — particularly about the risk of leaking information of potential use to attackers through social media, the BMJ advised.
Smart devices
Wearable and implantable smart medical devices are a proven cybersecurity risk. These technologies certainly offer better analysis, assisting diagnosis of medical conditions while aiding independent living, but mistakes made in securing such medtech have exposed vulnerable users to potential attack.
A seminal moment was the late Barnaby Jack’s hacking of an insulin pump in 2011. This attack over Bluetooth had a maximum range of approximately 300 meters.
Since then, security researchers at Pen Test Partners have found “closed loop” insulin trial data on the public internet.
“In one case, we could have modified the readings taken by the body-worn continuous glucose monitor and automatically, remotely administered a fatal dose of insulin to around 3,000 users in the trial,” Ken Munro, managing director of Pen Test Partners, tells CSO. “Fortunately, the vendor involved responded very quickly to our report and had the system secured the same day.”
Munro adds: “Other connected medtech devices Pen Test Partners have found security issues with include cranial stimulators, dosing pumps, and medical robots, among many others. Fortunately, the smart devices threat has been recognized and regulators are starting to take action.”
For example, the US Food & Drug Administration (FDA) introduced FD&C 524b in 2023 to drive cybersecurity in connected medical devices.
Generative AI
As healthcare staff adopt generative AI, the risk of leaking sensitive information through prompts and documents has grown.
Regulated data, such as patient records and medical information, is especially at risk, accounting for 89% of all data policy violations occurring in the context of gen AI usage, significantly higher than the cross-industry average of 31%, according to a 2026 study by Netskope.
Moreover, the Netskope report shows that healthcare organizations’ deployment and usage of internal AI tools, which require bespoke security guardrails, is accelerating. The proportion of healthcare workers using gen AI applications managed by their organization jumped from 18% to 67% in 2025, significantly ahead of cross-industry averages (26% to 62%), according to the study.
The need for bespoke security controls for AI systems is illustrated by research from Mindgard showing that the clinical AI tool Doctronic could be compromised to spread conspiracy theories or even manipulate prescription guidance.
View the full article
Cyber resilience has become a critical governance concern as organizations face increasingly complex and costly cyber threats. However, recent research reveals that the concept of cyber resilience remains inconsistently defined across regulatory frameworks and in some cases presents contradictory guidance to cross-sector and multinational organizations. This conceptual fragmentation poses a systemic risk for top management teams. Without a standardized definition, boards will struggle to determine what they should oversee, what should be measured, and how to evaluate overall organizational cyber resilience. This challenge is amplified by several external factors such as regulatory pressure, increasing public scrutiny, and the growing economic impact of cyber disruption.
What does the current research literature tell us?
In an effort to synthesize the views that currently exist on cyber resilience, I recently conducted a detailed literature review on 38 articles that examine how cyber resilience is defined and conceptualized in contexts that are relevant to boards of directors and executive teams. The review included academic research articles, industry white papers, and output from recognized cyber resilience working groups.
The goal was to assess whether existing definitions and conceptualizations of cyber resilience demonstrate sufficient alignment to support a generalized definition of cyber resilience relevant to boards and, over time, support meaningful measurement models. The results are clear: The field of cyber resilience is still in its infancy, and the understanding of cyber resilience concepts is far from consistent. While many cyber resilience concepts are intuitively understood within the technical ranks, cyber resilience must be put into specific contexts of business outcomes when discussing with Boards and Executive teams.
Following is a short synthesis of the literature review findings, focusing on areas of convergence and divergence across the body of literature.
Where the literature converges
Organizational outcomes vs. policy and controls
It’s consistently agreed that cyber resilience should be tied to organizational outcomes rather than technical controls and policies. Rather than focusing on metrics such as mean time to detection or number of security controls, organizational cyber resilience needs to evaluate levels of business continuity, preservation of stakeholder confidence and financial stabilization in the wake of major disruptions. To do so demands that cyber resilience be treated as a strategic priority that is integrated into organizational governance and business strategies.
Resilience is much broader than preparedness
While some articles framed risk prevention and protection in the context of resilience, business continuity and recovery featured prominently as important resilience themes in virtually every article reviewed. This demonstrates a common belief that disruption is inevitable and that resilience is demonstrated through rapid response and recovery rather than preparedness alone. Some articles went as far as to advocate for treating cybersecurity and cyber resilience as separate concepts to help address ambiguities.
Cyber resilience is a leadership responsibility
Cyber resilience is increasingly framed as a leadership responsibility, with the associated governance identified as one of the top governance challenges currently faced by boards of directors. Many sources explicitly position boards as being accountable for resilience outcomes, with some articles stressing the need to assign responsibility to a single officer. This is clearly driven by governmental regulations where, in some countries, Board responsibility for cyber resilience outcomes is explicitly identified. Articles also highlight the importance of senior leadership in fostering a culture of cybersecurity awareness and resilience across the entire organization.
Industry context
The review also investigated differences in definitions and conceptualizations of cyber resilience across industries, and while differences do exist, they are exclusively a reflection of prioritization and operating environment, not fundamentally different views of resilience. For example, articles that focused on the financial services sector emphasized regulatory compliance and systemic stability of the ecosystem, while those that focused on the energy and industrial sectors make it clear that the priority is on operational continuity and safety. While these differences might influence how resilience is implemented and measured, the underlying concepts of cyber resilience remain consistent across industries.
Diverging views
Conceptual framing: What exactly is cyber resilience?
One area where current cyber resilience literature diverges is in the overall framing or construct of cyber resilience. In many cases, cyber resilience is framed as a component of cybersecurity, while others consider cybersecurity a precursor to cyber resilience. In many of the reviewed articles, terms such as risk management, cybersecurity preparedness and cyber resilience were used interchangeably. Approximately 30% of articles framed cyber resilience as a very distinct construct, differentiating it from general cybersecurity concepts. These articles exclusively positioned cyber resilience as a broader strategic governance construct.
Scope of cyber resilience: Where does it start and where does it end?
The most glaring area of debate is where, in the lifecycle of a cyber crisis, responsibility for cyber resilience starts and ends. For example, as previously noted, it’s often unclear whether preparedness is part of resilience or a precursor to it. Nearly half of the research articles reviewed consider risk analysis and preparedness separate from resilience. In such cases resilience is limited to response, recovery and adaptation. Essentially, separating what happens before a crisis from what happens during and after a crisis.
This separation makes sense as cyber risk preparedness is generally measured through policies and controls, while crisis response, recovery and adaptation require much different measures. Still, slightly more than 50% of the articles view cyber resilience as an overarching construct with responsibility for all aspects from anticipating risk through to recovery and adaptation. This view also makes sense and is one that is more relevant to boards and executive teams given their broad fiduciary responsibilities to protect the company and its shareholders.
Cyber crime, the catalyst for cyber resilience: What about unintentional disruptions?
Similar to the previous area of debate is how broadly causes of disruptive events are defined in the context of cyber resilience. While cybercrime features prominently in the literature and remains the dominant driver of cyber risk, many definitions intentionally avoid tying resilience to a single type of threat. In fact, of the 38 articles reviewed, only two specifically linked their definitions of cyber resilience to events resulting from cybercrime. This was a surprising realization, but it is encouraging when considering the number of recent highly visible unintentional disruptions that have caused massive outages. Surely, cyber resilience governance needs to focus on an organization’s capacity to absorb and recover from any type of cyber disruption — intentional or not.
Governmental regulation: An enabler or a constraint?
Regulatory frameworks for cyber resilience vary significantly across geographies and industries, creating challenges for standardization. Cyber regulations within the financial services sector are quite different to those in the energy sector, which are different again than those in the consumer technology sector. Several of the articles reviewed highlighted the complexity of navigating the sea of regulations and regulators. Relevant regulators include CISA, FTC and SEC in the United States, more than 10 sector specific regulators in the UK, individual member states within the EU, and many more.
There is some disagreement on the benefits of regulation, but the majority opinion from this review is that for organizations, particularly multinationals, the complexity has become unmanageable. It is also noted that for small and medium enterprises, compliance with regulations can provide a false sense of security, as it can be tempting to assume compliance equates to resilience.
Implications for boards and executive teams
Boards and executive teams are increasingly expected, and in some cases required by law, to bear responsibility for overall cyber risk and resilience outcomes. This responsibility far exceeds historical expectations of simply approving security investments or compliance activities. To do this effectively, it’s imperative that a clear definition and scope of cyber resilience be clearly communicated in business terms to the board and executive team. For such audiences, cyber resilience must be framed in terms of operational impact, financial exposure, and organizational continuity rather than technical indicators alone.
As economies become more integrated and interdependent, their exposure to and impact from cyber disruptions grow exponentially. As such, the ability of individual organizations to withstand and quickly recover from cyber disruption is not just a measure of organizational cyber resilience; it also contributes to our overall global economic resilience. In this sense, cyber resilience should be considered one component of a larger resilience agenda, alongside financial, operational, and supply-chain resilience. Clarifying the scope of cyber resilience at the board level is therefore not only a governance issue, but also a foundation for understanding overall resilience in increasingly digital economies.
This article is published as part of the Foundry Expert Contributor Network.
Want to join?
View the full article
Microsoft on Tuesday released updates to address a record 169 security flaws across its product portfolio, including one vulnerability that has been actively exploited in the wild. Of these 169 vulnerabilities, 157 are rated Important, eight are rated Critical, three are rated Moderate, and one is rated Low in severity. Ninety-three of the flaws areView the full article
Built by a veteran security team and led by a former Google and Mandiant executive, Mallory delivers intelligence that drives action for enterprise security teams.
Mallory is launching a AI-native threat intelligence platform, purpose-built to answer the questions CISOs and their teams are asking every day:
What are the real threat vectors for our organization? What’s actually exploitable in our environment right now? What should we proactively fix? The platform monitors thousands of threat sources, contextualizes them against a user’s actual attack surface, and puts that intelligence to work across hunt, detection, and exposure management use cases. One platform. Answers, not alerts.
Modern teams receive more signals than ever and have structured their SOC around dealing with the deluge. But they’re still chasing alerts and struggling to get proactive, let alone work with business partners to close gaps before they’re exploited.
Mallory connects into the user’s existing tools and controls. When a new vulnerability or exposure surfaces, it doesn’t just flag it — it tracks who is exploiting it, where, and how, then determines whether the organization is actually at risk and what to do about it.
“Attackers are AI-enabled now, moving faster and with more capability. Defenders need to be too. Security teams don’t need more alerts. They need answers: what can attackers do, are our controls stopping them, and what’s exploitable right now,” says Mallory founder and CEO Jonathan Cran.
The result isn’t another feed or dashboard. It’s a prioritized set of evidence-based cases grounded in real threat intel, mapped to the user’s environment and ready for action.
“When a new alert makes the news, I need to know within minutes if we are impacted. Mallory delivers the context needed to investigate at AI speed,” says John Sapp, CISO of Texas Mutual Insurance.
Flexibility to build is critical in today’s cybersecurity ecosystem. Mallory is built by veteran security practitioners for security teams, with native support for Claude Code, MCP, API, and its own modern UI. Teams can integrate, automate, and extend on their terms.
Mallory also announced a seed investment led by Decibel Partners, with participation from Live Oak Venture Partners and a cadre of industry leaders from organizations including Google, Robinhood, Cisco, Fastly, and GreyNoise.
“Threat intelligence was built for an era where we would be able to process information at human speed. With the introduction of agents on the adversarial side, we no longer have data intel problem but rather a context and reasoning problem. Jonathan and the Mallory team are changing that by connecting real-time threat activity to an organization’s environment and processing it for relevance at agentic speed. ” says Dan Nguyen-Huu, partner at Decibel.
Mallory is available immediately as a SaaS platform with integrations across existing security tools. Users can start a 30-day free trial at mallory.ai/platform.
About Mallory
Mallory is the AI-native threat intelligence platform for cyber defenders. It monitors global adversary activity, contextualizes threats against assets and controls, and delivers prioritized, evidence-based answers. Teams focus on what’s real and act faster. Users can learn more at mallory.ai.
Contact
Marketing
Chris Tilton
Mallory
[email protected]
View the full article
OpenAI on Tuesday unveiled GPT-5.4-Cyber, a variant of its latest flagship model, GPT‑5.4, that's specifically optimized for defensive cybersecurity use cases, days after rival Anthropic unveiled its own frontier model, Mythos. "The progressive use of AI accelerates defenders – those responsible for keeping systems, data, and users safe – enabling them to find and fix problemsView the full article
Miljan Zivkovic | shutterstock.com
Die zunehmende Abhängigkeit von IT-Dienstleistern und Software von Drittanbietern vergrößert die Angriffsfläche von Unternehmen erheblich. Das wird auch durch zahlreiche Cyberattacken immer wieder unterstrichen. Zwar lassen sich die Risiken in Zusammenhang mit Third-Party-Anbietern nicht gänzlich beseitigen, aber durchaus reduzieren. Dabei sollten Sicherheitsentscheider eine zentrale Rolle spielen, wie Randy Gross, CISO bei CompTIA, erklärt: “CISOs sind in der einzigartigen Lage, den gesamten Geschäftsprozess zu überblicken – Datenflüsse, Abhängigkeiten und nachgelagerte Auswirkungen. Dennoch nutzen viele Unternehmen diese Perspektive noch immer nicht, um Risiken durch Dritte neu zu bewerten.”
Insbesondere, wenn Verträge auf Ebene von Geschäftseinheiten verhandelt werden oder unterhalb der finanziellen Genehmigungsschwellen liegen, bleiben Sicherheitschefs jedoch oftmals außen vor. Das beobachtet auch Melissa Ventrone, Leiterin der Abteilung für Cybersicherheit bei der Anwaltskanzlei Clark Hill: “In vielen Unternehmen werden Sicherheitsverantwortliche erst nach Vertragsabschluss hinzugezogen. Oder wenn ein Sicherheitsproblem bereits aufgetreten ist.”
Tatsächlich sollten CISOs an dieser Stelle als pragmatische Technologieberater fungieren, die wichtige Informationen einholen, für deren Bewertung sie besonders qualifiziert sind.
13 Fragen, die Sie Drittanbietern stellen sollten
Die folgenden Fragen unterstützen CISOs und Sicherheitsentscheider dabei, das in der Praxis umzusetzen.
1. Welche Nachweise für angemessene Sicherheitskontrollen können Sie erbringen?
Laut Juan Pablo Perez-Etchegoyen, CTO beim Security-Anbieter Onapsis, zählen folgende Nachweise dabei zu den gängigsten:
SOC 2 Typ II (gilt als Goldstandard für Auditierungen von IT- und Cloud-Dienstleistern), ISO/IEC 27001, Cloud Security Alliance STAR (speziell für Cloud-Anbieter, kombiniert ISO 27001 mit einer Kontrollmatrix für Cloud-bezogene Risiken), sowie branchenspezifische Zertifizierungen (zum Beispiel HIPAA/HITRUST für den Umgang mit Gesundheitsdaten oder PCI DSS für die Verarbeitung von Kreditkartendaten). 2. Wie werden diese Kontrollen aktualisiert und wie werden wesentliche Änderungen kommuniziert?
Anwältin Ventrone empfiehlt zudem,potenzielle IT-Partner mit einem detaillierten Due-Diligence-Fragebogen zu konfrontieren. Darüber hinaus empfiehlt die Rechtsexpertin, spezifische Aspekte vertraglich zu verankern: “Drittanbietern sollte es mindestens untersagt sein, Sicherheitskontrollen zu verändern, die das Schutzniveau oder die Ausfallsicherheit Ihrer Systeme und Daten beeinträchtigen würden.”
3. Wer ist in Ihrem Team für die Identity Posture zuständig und wie erkennt derjenige Anfragen, die auf Social Engineering zurückzuführen sind?
Welche Form von Zugriff das Team des Drittanbieters auf Kundensysteme und -daten hat und wie dieser segmentiert und abgesichert ist, sollten Sicherheitsentscheider nach Meinung von Casey Corcoran, Field CISO beim Managed-Service-Anbieter Stratascale, unbedingt erfragen. “Achten Sie darauf, dass dieser Zugriff protokolliert sowie überwacht wird – und bei Bedarf sofort widerrufen werden kann.”
Zudem sollten Sicherheitsverantwortliche dabei die richtigen Aspekte ins Auge fassen, wie John Alford, CSO bei der Unternehmensberatung TeraType, betont: “Viele Kunden konzentrieren sich auf Firewalls, Endpunkt-Agenten und MFA – übersehen dabei aber die Trust-Pfade, die Angreifer bevorzugt nutzen: Helpdesk-Workflows, OAuth-Integrationen, Lieferanten-Support-Portale und Automatisierungs-Konnektoren.”
Alford empfiehlt seinen Berufskollegen außerdem, auf streng definierte Rollenbereiche, mehrstufige Verifizierungen sowie Approval Chains für den Reset von Anmeldedaten zu achten. “Ist nichts davon vorhanden, deutet das auf blinde Flecken hin, die sich nachträglich nicht beseitigen lassen.”
4. Wie lassen sich Ihre Workflows verifizieren? Können Sie Nachweise über deren Wirksamkeit erbringen?
Viele Unternehmen unterschätzen, wie viel operatives Vertrauen sie wirklich an Anbieter abgeben. Deswegen sollten Drittanbieter nicht nur Richtlinien-Dokumente vorweisen können, sondern auch Workflow Maps, Execution-Protokolle und Testing-Belege. “Die größten Lücken treten regelmäßig an den Stellen zutage, die vermeintlich sicher sind. Ich habe schon erlebt, wie gestandene Unternehmen mit ausgeprägten Standards und Kontrollmaßnahmen an Problemen gescheitert sind, für die ausschließlich die Workflows ihres Third-Party-Anbieters verantwortlich waren”, plaudert Alford aus dem Nähkästchen.
Risikobewertungen sollten sich seiner Meinung nach deshalb nicht bloß auf auf Server und Netzwerke konzentrieren, sondern auch auf Identitäts-Workflows und manuell gesteuerte Prozesse: “Wenn man den Blickwinkel erweitert, entdeckt man unter Umständen Kontrollmaßnahmen, die lediglich auf dem Papier gut aussehen.”
5. Welche Rolle spielt unabhängiges Testing bei Ihnen und wie oft wird das eingesetzt?
Geht es um Security-Tests und -bewertungen bei IT-Partnern, sollten CISOs Wert darauflegen, dass diese von unabhängigen Dritten durchgeführt werden, meint Rechtsexpertin Ventrone. “Das sollte mindestens einmal pro Jahr stattfinden – und bei wesentlichen Änderungen an Netzwerk, Infrastruktur oder Sicherheitskontrollmaßnahmen. Die Zusammenfassungen von Schwachstellen-Scans, Penetrationstests und Audits sollten Sie sich ebenfalls zeigen lassen.”
Danny Jenkins, CEO beim Security-Anbieter ThreatLocker, stellt insbesondere auf die Frequenz dieser Überprüfungen ab: “Bedrohungen entwickeln sich ständig weiter. Eine jährliche Prüfung reicht deshalb nicht aus. Sämtliche Systeme sollten regelmäßig Penetrationstests unterzogen und optimiert werden.”
6. Können Sie sämtliche OAuth-Integrationen und API-Beziehungen in Ihrem Service auflisten und erklären, wie diese definiert, überwacht und widerrufen werden?
OAuth-Integrationen werden nach Einschätzung von Teratype-CSO Alford allzu oft als harmlose Annehmlichkeiten behandelt – statt als High-Privilege-Kanäle: “In Wirklichkeit funktionieren sie wie ein Netzwerk aus vergessenen Tunneln. Sie bieten eine Möglichkeit, das Eingangstor vollständig zu umgehen und verbinden Systeme tief im Inneren der Umgebung.”
Unternehmen sollten ihre Drittanbieter deshalb auffordern, ein Token-Inventar inklusive Minimal Scopes, endlichen Laufzeiten sowie einer Möglichkeit zum Behavioral Monitoring bereitzustellen, so Alford. Permanent gültige Token sieht der Experte hingegen als Warnsignal, das auf ein erhöhtes Risiko hindeutet.
7. Wie sehen Ihre vertraglichen und operativen Pflichten aus, wenn ein Angreifer Ihre Prozesse missbraucht, ohne dabei in Systeme einzudringen?
Wenn Drittanbieter Passwörter zurücksetzen oder OAuth-Integrationen managen können, wird der Vertrag zu einem Kontrolldokument. Dieses definiert, wie das Risiko geteilt wird und welche Nachweise der Kunde verlangen kann. An dieser Stelle ist es besonders wichtig, Sicherheitsentscheider in die Verhandlungen mit Third-Party-Anbietern einzubeziehen, wie Alford betont: “Ohne die Beteiligung des CISO bleiben Vertragsklauseln in der Regel eher nachteilig ausgestaltet. Das liegt daran, dass der Fokus ohne Security-Perspektive vor allem auf der Verfügbarkeit liegt – und nicht so sehr auf der Sicherheit. Als Kunde sollten Sie darauf bestehen, dass sich die Pflichten des Anbieters nicht nur auf kompromittierte Systeme, sondern auch kompromittierte Prozesse erstrecken.”
8. Welche Kontrollmaßnahmen kommen zum Einsatz, um die Aktivitäten Ihrer Mitarbeiter in unserer Umgebung zu kontrollieren? Und wie erkennen wir Verhalten, das von der Norm abweicht?
“Moderne Angriffskampagnen machen sich Vertrauensbeziehungen und weiche Betriebsprozesse zunutze. Die Gefahr lauert dabei oft dort, wo sie niemand erwartet – beispielsweise Helpdesks”, warnt Alford. Die Aktivitäten der Belegschaft von Drittanbietern zu überwachen sei daher erfolgsentscheidend. Der Sicherheitsprofi empfiehlt deshalb, darauf zu bestehen, dass der Partner Sessions aufzeichnet, Echzeit-Alarmmeldungen zur Verfügung stellt und Aufgaben strikt getrennt werden.
9. Wie isolieren Sie unsere Assets und Daten von denen anderer Kunden?
Mit Blick auf potenzielle Third-Party-Anbieter sollten CISOs zudem auf eine klare Architektur und konkrete Maßnahmen achten, die Schaden begrenzen können. Dabei spielt auch eine Rolle, wie der Drittanbieter Risiken in seinen eigenen Lieferketten managt. “IT-Partner sollten über ein robustes Vendor-Management-Programm verfügen und ihre eigenen Dienstleister einer angemessenen Due-Diligence-Prüfung unterziehen”, rät Ventrone.
10. Wie schnell werden wir über Sicherheitsvorfälle informiert, die sich auf unsere Daten oder Systeme auswirken?
Von IT-Partnern über potenzielle Sicherheitsvorfälle informiert zu werden, sollte selbstverständlich sein. Dabei sollte jedoch auch eine Rolle spielen, wie zeitnah das erfolgt. Stratascale-Field-CISO Corcoran empfiehlt diesbezüglich: “Der Vertrag sollte eine Meldung des Drittanbieters innerhalb von 24 bis 72 Stunden garantieren. Darüber hinaus sollten Security-Verantwortliche auch auf einen getesteten Incident-Response-Plan sowie weitere vertraglich verankerte Verantwortlichkeiten achten.”  
Auch Alford sieht bei diesem Punkt kein Potenzial für Kompromisse – Drittanbieter müssten ihren Kunden ausreichend Informationen zur Verfügung stellen, damit diese ihre eigenen Threat-Analysen fahren können: “Geschieht das nicht, können sich Kundenunternehmen lediglich noch auf die Detection- und Reporting-Funktion des Hosting-Anbieters stützen.”
11. Wie identifizieren, priorisieren und beheben Sie Schwachstellen?
Da nicht wenige Cyberattacken auf bereits bekannte Schwachstellen abzielen, gilt es bei der Evaluierung von Drittanbietern auch auf Patch-Richtlinien und Remediation-Fristen zu achten. Onapsis-CTO Perez-Etchegoyen klärt über die Risiken in diesem Zusammenhang auf: “Langsame Patch-Zyklen können zu Lieferkettenunterbrechungen, betrieblichen Problemen und manchmal auch zur Insolvenz führen.”
Ventrone führt an dieser Stelle das Beispiel eines Unternehmens an, das sein Firewall-Management an einen Drittanbieter ausgelagert hat: “Nachdem eine Schwachstelle in der Firewall ausgenutzt worden war, stellte der Partner schließlich die anfällige Version wieder her – was zu einer zweiten Kompromittierung führte. Um es salopp zu sagen: So etwas kann man sich nicht ausdenken”, konstatiert die Anwältin.
12. Verfügen Sie über eine Cyberversicherung, die mögliche Auswirkungen auf sämtliche Ihrer Kunden abdeckt?
Laut Joshua Wright, Faculty Fellow beim SANS Institute, werden die Attacken auf SaaS-Anbieter in Zukunft weiter steigen – und damit auch die nachgelagerten Risiken: “Wird ein solcher Drittanbieter kompromittiert, entstehen diverse Möglichkeiten für Folgeangriffe – etwa mit Ransomware.”
Ventrone empfiehlt CISOs deshalb, in Verhandlungen mit Drittanbietern auch sicherzustellen, dass deren Cyberversicherungspolice nicht nur das eigene Unternehmen abdeckt, sondern auch den gesamten Impact eines Vorfalls, bei dem mehrere Kunden betroffen sind.  
13. Können wir Ihre Prozesse testen?
SANS-Experte Wright hält darüber hinaus auch Nachweise für Testing und Monitoring für unerlässlich – etwa bezogen auf Penetrationstests, Security Monitoring oder Threat Hunting. Alford empfiehlt allerdings, noch einen Schritt weiter zu gehen: “Prozesstests unter Einbeziehung realistischer Szenarien können aufdecken, wo tatsächlich Risiken bestehen. Das gibt Ihnen zudem die Möglichkeit, Kontrollen zu entwickeln, die der Denkweise von Angreifern entsprechen und nicht dem, was in der Dokumentation steht.” (fm)
View the full article
In 2026, enterprise developers are building and deploying the first generation of powerful, increasingly autonomous AI agents at incredible speed. Now comes the hard part: working out how to secure them.
Vendors in the space are facing multiple challenges. To begin with, traditional identity and access management (IAM) tools were never designed to secure anything as complex as agentic AI. In addition, the number of agents, both those sanctioned by the enterprise and the undocumented ‘shadow’ agents created by a new generation of powerful tools that barely existed a year ago, is increasing at unprecedented speed. And now it has started to dawn on organizations that this risks leaving yawning governance and security gaps whose weaknesses could one day return to haunt their creators.
While a growing list of companies, including large cloud platforms such as Okta, Ping Identity, and Microsoft’s Entra ID, is vying to fill the vacuum, a smaller competitor, Sweden’s Curity, argues that agents can’t be secured using traditional IAM. Instead, it is offering a different approach to the problem: This week, it announced Access Intelligence, an extension to its existing API identity and access management (IAM) platform, Identity Server.
The problem it addresses is that traditional IAM tools assume that applications are being accessed by human users or machine identities, governed by a one-time authentication process. But agents, which assume long chains of actions conducted at incredible speed, don’t work like this. Instead, access becomes ephemeral, complex, and non-deterministic, which is to say, hugely unpredictable. Lock them down too much and they stop working; let them run free, and weak security follows in their wake.
Runtime enforcement
Curity’s approach is to treat agents as a special type of application. Like applications, agents call APIs, MCP servers, and each other, and are credentialed using OAuth tokens. Through a feature called Token Intelligence, Curity extends the role of OAuth tokens to not simply permit access, but to carry information on the agent’s purpose and intent. In Curity’s scheme, an agent can only access resources based on that purpose.
Instead of using static, pre-granted permissions, agent access is granted at runtime, on-the-fly. Each requested action generates a separate token that describes the access it needs. When an agent starts a new task, it needs a new token specifying a new set of permissions. If necessary, human authorization can be required when an agent is trying to perform a high-risk action such as transferring funds.
“Curity has always been application-centric,” said Cofounder and CTO Jacob Ideskog. “Our focus has always been on how we broker access.”
Multiple approaches to agent security
Today, agent security falls into one of several camps, which include increasingly inadequate inline approaches such as API gateways and web application firewalls (WAFs), and out-of-band analysis systems that infer intent by analyzing agent behavior against a baseline.
Curity’s Access Intelligence, by contrast, is a self-hosted microservice that acts as a glorified IAM layer through which every agent request must pass. “Because we let an agent do something now doesn’t mean we should be allowing it to do this a minute later,” Ideskog explained.
Access Intelligence also uses Identity Server’s centralized token validation to ensure that developers can fire up agents or APIs without registering them. If they lack this validation, agents are isolated from real-world actions.
Nothing does the whole job
The appearance of systems such as Access Intelligence is good news for enterprises. It indicates that vendors are starting to address the problem of agent security, often by extending existing API security platforms. But that still leaves open the question of which approach to take.
Ideskog believes it would be a mistake to see the different approaches as mutually exclusive. Curity’s Access Intelligence can be used in combination with other layers of agent security, he emphasized. In short, no one solution can do the whole job.
“Up to this point, the IAM industry has focused on the identity part. But the real question is the access. Enterprises are asking their privilege access management (PAM) vendors how they’re going to deal with this [agent security] and I don’t think the PAM vendors have good answers yet,” he said.
View the full article
Microsoft increased prices for all of its Surface PCs this week, with most models priced hundreds of dollars higher than they were when launching. Windows Central highlighted the increases, which now see Microsoft's mid-range models priced above $1,000 and flagship models priced starting at $1,500.


A Microsoft spokesperson said the price increase was due to "recent increases in memory and component costs."

Microsoft's 12-inch Surface Pro, which was its cheapest modern PC at $799, is now priced starting at $1,049. The flagship 512GB 13-inch Surface Pro is $1,499, up from $1,199 when it launched in 2024 (Microsoft also discontinued a $999 256GB configuration). The 13-inch Surface Laptop went from an $899 starting price to a $1,149 starting price, while the 13.8-inch model went from $999 to $1,499 and the 15-inch model went from $1,299 to $1,599.

The 13-inch Surface Pro and the 13.8-inch and 15-inch Surface Laptop models originally launched in 2024, and Microsoft did increase prices for them in 2025, so this is the second price increase. The 13-inch Surface Laptop and the two Surface Pro models that have seen a $300 price increase launched in 2025.

Microsoft's 13.8-inch Surface Laptop 7 with 16GB RAM and 256GB of storage used to be $100 cheaper than the 256GB M4 MacBook Air, but now it's $400 more than the 512GB M5 ‌MacBook Air‌. Apple increased ‌MacBook Air‌ pricing from $999 to $1,099 with the M5 upgrade, but Apple's hike came with more base SSD storage. The Surface Laptop 7 is the laptop that Microsoft says is "faster than a ‌MacBook Air‌ M4."

Prices have increased for all Surface Pro and Surface Laptop models, from entry-level to high-end. Microsoft's PCs are now more expensive than their Mac equivalents, which is good news for Apple. The high-end Surface Laptop 7 with 64GB RAM and a 1TB SSD is $3,649, which is more expensive than the 16-inch $3,300 M5 Pro MacBook Pro with 64GB RAM and a 1TB SSD. Apple's M5 Pro chip also far outperforms the Snapdragon X Elite.

Windows Central says Microsoft has new Surface PCs coming later this year, which are also expected to have the same higher prices.

Microsoft's decision to increase PC prices comes as Samsung also raised prices for some of its smartphone models and all of its U.S. tablet offerings.

Both Microsoft and Samsung are responding to increased costs caused by global memory shortages. Chip makers are prioritizing memory for AI data centers, and there is little manufacturing capacity left for consumer devices.Tag: Microsoft
This article, "Microsoft Raises Prices for All Surface PCs, Making Them More Expensive Than Equivalent Macs" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Apple removed scam app Freecash from the App Store this week after the app spent months harvesting data from iPhone users, reports TechCrunch.


Freecash reached the number two spot on the U.S. ‌App Store‌ charts in January after being heavily marketed on TikTok. It promised users up to $35 per hour for watching TikTok content, but it was collecting swaths of user data. Back in January, Wired covered Freecash's deceptive marketing, and MalwareBytes pointed out that the app was gathering data like race, religion, health, and biometrics, with extra data harvested through mobile games that Freecash pushed users to install.

Users tricked into downloading Freecash with the promise of free cash found that they could not earn money by using TikTok, but instead were able to earn tiny amounts of cash by playing games like Monopoly Go and Disney Solitaire. The goal was to push users to make in-app purchases or watch paid ads in the apps. Freecash advertised itself as a platform for matching game developers with users likely to spend money in their games.

After the Wired report, TikTok pulled the Freecash ads, but Apple did not take action to remove the app. Freecash stayed in the ‌App Store‌ until TechCrunch contacted Apple on Monday, which is when Apple removed the app from the ‌App Store‌. Apple said Freecash violated its guidelines prohibiting scam practices and misleading marketing.

Freecash parent company Almedia denied using deceptive marketing techniques and said it was in compliance with Apple's ‌App Store‌ rules.

Freecash was downloaded by 5.5 million people across the Apple ‌App Store‌ and Google Play in January 2026, and it has remained high on the ‌App Store‌ charts since then. In addition to using misleading TikTok ads, the app appears to have used bots and fake ratings to drive traffic. The app's developers may have also acquired an existing ‌App Store‌ app to get around Apple's app review system, as the Freecash app was banned at one point in 2024 before an existing ‌App Store‌ app was renamed Freecash and updated with the same functionality.

TechCrunch's full report has more information on Freecash and the scammy tactics the company behind the app used to lure in users.Tag: App Store
This article, "Apple Removes Freecash App From App Store After Months of Data Harvesting" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
SwitchBot makes some of the most affordable HomeKit-compatible robot vacuums on the market. I've been testing the SwitchBot S20 and the K11+ for the last several months to see how they measure up to some of the more expensive models that I've reviewed.


SwitchBot S20

The MSRP for the S20 is $799, but SwitchBot runs sales often that drop the price to under $500, so it's competitive with some of the more affordable robot vacuums that don't have Matter integration.

I've noticed some clear differences between the SwitchBot vacuums and higher-end vacuums I've tested, but the S20 has impressed me with its cleaning ability. It is a vacuum and a mop, so it can vacuum up crumbs and mop the floors at the same time.


For vacuuming, it has a 10,000Pa suction, but I've never been able to tell a real difference in suction power between the vacuums I've tested. They're all able to do a good job with dust, fur, dirt, and other debris on the floor. There is an anti-tangle system, and I've never seen the SwitchBot S20 have a problem with hair or fur, and there are rubber ridges on the vacuum brush that seem to help it pick up lightweight messes. A brush on the side sweeps along baseboards, and while I don't think it's as good at getting into small crevices as the Roborock vacuum I have, it does a passable job. You might have to do a perimeter sweep with a regular vacuum every once in a while.

The S20 has a roller mop, and that's my favorite robot vacuum mopping design. A lot of companies use rotating mopping pads, but the roller gets washed continually as the robot cleans, so it's not spreading a mess around. It also just feels more hygienic to me because it's scrubbing grime off of the mop as it goes. No robot vacuum is able to get crusty, dried-on stains off in one pass, but the S20 did a good job on dried ketchup with two passes. Robot vacuums aren't ideal for deep cleaning, but with multiple cleans per week or even every day, my floor is cleaner than when I do it manually.


I have wood flooring and tile, but no carpet, so I can't test carpet extensively. I have a few larger-sized rugs that it has done fine vacuuming, and it's done no damage to my wood floors.

All robot vacuums have a base station, and the S20 is no exception. In fact, it has the option for a base station that hooks up to your plumbing so it can automatically empty waste water and refill with clean water. For a review, I didn't want to hook it up to my house permanently so I opted for tanks that I refill and empty manually, but it's nice to have the option. The SwitchBot S20 base station has a simple, unobtrusive design, and it's not overly large.


There's a 2.7L clean water tank and a 2.5L dirty water tank, which I have to refill/empty around once a week depending on how often the robot runs. The base station also has a spot for adding floor cleaner, and there are cleaners that are for robot vacuums. The tanks are easy to access, simple to open, and quick to refill or empty.

Dust and dirt are collected in a bag that needs to be replaced every three months, so that is an ongoing expense. You'll also need to do some light cleaning and maintenance of the robot vacuum every so often because the mop and other components don't last forever, but that usually only needs to be done once a year or so. The base station uses hot air (50°C) to dry the mop after cleaning, and that keeps it from growing mold or smelling mildewy.

Those are all the good things about the SwitchBot, and now I'll go over the not-so-good and the bad. The base station has a thin plastic mat to protect the floor from the wheels, which SwitchBot wants you to attach to the floor with adhesive. I am not going to attach plastic to my floor, and I was disappointed there wasn't a snap-in floor protector like most robot vacuums have for the base station. Without the adhesive, the plastic moves around and is useless, so I just deal with some drips and dirt on my kitchen floor.


Compared to Ecovacs and Roborock vacuums I've tested, the S20 is loud. I would not be able to sleep or work through its vacuuming noise, though it does have a mode to reduce suction and make it quieter. It's still not quiet enough. It sounds like a vacuum when the vacuum is running, but it's not the worst tradeoff for a lower price.

The S20 has modes for vacuuming, vacuuming and then mopping, or vacuuming and mopping at the same time, but if you just want it to mop, SwitchBot doesn't have an option for that.


As far as the AI goes, it's not as good as some of the other vacuums I've used, and in some ways that's a good thing. It doesn't get stuck on my kitchen rugs because it just flat out ignores them. If one is stuck on the wheel, it drags the rug along until it gets unstuck. I don't necessarily mind, because I don't have to get up and save it. It's able to operate independently for the most part, and I don't have to intervene often.

It is able to avoid cords almost entirely, and I haven't had it suck up anything that it shouldn't except for a cat toy and a sock. The Ecovacs robot I tested would sometimes err too far on the side of caution, misidentifying objects and staying away from them for a less thorough clean. The S20 cares less, and that could be an issue if you have a lot of items on the floor.

The AI mapping isn't as capable as some more expensive robots, but the S20 was able to identify every room in my house and it navigates them well for the most part. It is not great at thresholds, especially taller thresholds. It gets stuck in my bathroom, and instead of realizing it's stuck and alerting me, it will keep trying to get out until its battery is exhausted.


I can edit maps to create no-go areas, label rooms and objects, and make other edits to make sure that it's only cleaning where I want it to clean, and the edits are a must with the SwitchBot vacuums.

The S20 can go for around 100 minutes before it needs to charge (in vacuum and mopping mode), and can clean approximately 1,000 square feet in my house before that point. It isn't able to do my entire house on a single charge, but in-app scheduling lets me have it clean a room or an area a day on a cycle. The battery lasts closer to three hours in vacuum-only mode.

SwitchBot K11+

The SwitchBot K11+ is a much smaller, vacuum-only robot. It's best for small spaces and it's a robot I'd consider in an apartment. With three attempts, I wasn't able to get it to accurately map my entire house, and the AI seems to be limited.


I can isolate it in a room and get an accurate map, so it's okay in a smaller area, but it's still not particularly intelligent. Like the S20, if it can't get to an area, it doesn't give up and move on. It continues to try to get there until it dies and I have to go hunt it down, and that's inconvenient when I'm not home.


The suction is decent at 6,000Pa per SwitchBot, but it doesn't pick up as much as quickly as the S20. What I like best about the K11+ is the small size. It's able to get in smaller nooks and crannies than bigger robot vacuums, which makes it ideal for small spaces.


There is technically a feature where you can attach a Swiffer-style mop to the K11+ for a mopping feature, but that just seems like much more of a hassle than quickly mopping the floor myself.

Matter Integration

With an Apple Matter hub (Apple TV or HomePod), SwitchBot vacuums connect to the Home app. Basic functions can be controlled through the Home app or through Siri, and I've come around to ‌Siri‌ integration as a useful feature.


I wasn't impressed with the limited robot vacuum controls in the Home app to begin with, but I can say things like "‌Siri‌, vacuum the kitchen" or "‌Siri‌, mop the dining room" to get a targeted clean when needed, and that's come in handy.

That's primarily what I use Matter for, but the Home app also supports automations and integration with other Apple products. You can have a setup where the robot vacuums when you leave home, so you never have to deal with the sound.

There is no situation where ‌Siri‌ or the Home app can be used to control a robot vacuum entirely without the need to access the dedicated SwitchBot app. Features like editing a map, updating firmware, viewing cleaning progress, troubleshooting, or checking estimates for replacement parts require the SwitchBot app, and that's true of any robot vacuum.

HomeKit integration is nice to have, but it is limited, and it's not a feature that I would choose one robot vacuum over another for.

Bottom Line

If you're looking for a robot vacuum and mop that does a good job cleaning and mopping and you don't want to spend a ton, the SwitchBot S20 is worth checking out. You'll need to deal with some frustrations, like loud noise, map editing, and rescuing it from being stuck, but it cleans well.

I would not recommend the smaller K11+ unless you have a small space. It's not meant for larger homes, but it is a good apartment vacuum if you won't miss mopping functionality.

How to Buy

The SwitchBot S20 is available from the SwitchBot website for $520 after a 35 percent discount (I've seen the price lower, so it's worth waiting for a bigger sale), while the SwitchBot K11+ is available for $220 after a 45 percent discount.

Note: SwitchBot provided MacRumors with an S20 and a K11+ for the purpose of this review. No other compensation was received.
This article, "SwitchBot S20 Review: A Budget-Friendly Robot Vacuum and Mop With Matter Support" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
We’re coming up on a year since launching Docker Hardened Images (DHI) this May, and crossing a milestone earlier this month made me stop and reflect on what we’ve actually been building.
Earlier this month, we crossed over 500k daily pulls of DHIs, and over 25k continuously patched OS level artifacts in our SLSA Level 3 pipeline. From the time we launched the free DHI Community tier at the end of last year, the catalog has now grown to 2,000+ hardened images, MCP servers, Helm charts, and ELS images. We continuously patch every artifact (across CVEs, distros, versions), so we’re now running over a million builds running regularly, and just getting started. Catalog coverage will jump again soon, as more Debian packages, ELS images, and newer artifact types are added.
But the numbers aren’t the interesting part. What matters is how we got here.
We chose the harder path, on purpose. Every product and engineering decision we made was consistently harder to build and operate, but better for developers and for the security of the ecosystem. We made hardened images free and open source. We built a multi-distro product, so adoption doesn’t mean migrating to a vendor’s proprietary OS. We build every system package from source for distros you already run. We ship a huge range of signed attestations with every image because that’s what independent verifiability actually requires.
Along the way, we also looked closely at how the rest of the industry approaches the same problems, and found patterns in patching timelines, SBOM completeness, and advisory coverage that are worth understanding before you evaluate any hardened image provider.
We made hardened images widely accessible so every team could raise their security baseline
We wanted to make a real dent in the security posture of the internet, and that meant making hardened images widely accessible. That is why we did not put our catalog behind a gated paywall, as was the industry norm, but freely available to every developer.
Building and sustaining a hardened image pipeline at this scale isn’t trivial. We know because we’ve been doing this for over a decade with Docker Official Images, freely for the community.
With the release of DHI Community under a permissive Apache 2.0 license, we raised the baseline for security across the ecosystem. Security should not be a premium feature. That kind of impact, at scale, is only possible because the foundation is open.
We built multi-distro so adoption is drop-in, and does not impose a migration tax on you
Some vendors in this space created an entirely new Linux distribution and called it “distroless,” which is a remarkable piece of branding for what is, in practice, a proprietary OS that your teams have never run, tested, or audited. Established Linux distributions like Debian and Alpine have a name for a package repository that only tracks the latest upstream version: they call it “unstable” or “edge,” not stable.
Docker doesn’t ship its own distribution, we harden the ones you already trust. That decision optimizes for your engineering reality, not ours. The hardened image that never gets adopted provides zero security value, full stop. 
With the Docker “multi-distro” approach, we support both Debian and Alpine today, with support for more distros to come. This is actually hard to do: the Debian and Alpine ecosystems don’t just differ in packaging; they diverge in libc, dependency trees, CVE streams, patch timing, and tooling. You are effectively maintaining parallel supply chains, each with its own nuances and security posture. Every hardened image in the DHI catalog is available in both Alpine and Debian, across both amd64 and arm64 architectures, which means we build, patch, and attest each combination independently, taking on that operational burden so you don’t have to.
We regularly speak with engineering teams who evaluate proprietary distributions from other vendors and run into the same wall: your existing internal expertise, tools, tests, and pipelines are built around Alpine or Debian.
Migrating to an unfamiliar, vendor-owned OS isn’t a security upgrade, it’s an adoption project and a material line item of cost, alongside the sticker price of the hardened images subscription itself. The vendor lock-in aspect goes without saying.
The migration effort means revalidating CI pipelines, retraining platform teams, auditing an entirely new package ecosystem, and working through compatibility gaps that surface weeks into a rollout. Several teams tell us they bought the migration story, spent months on it, and are still paying for images their engineers haven’t adopted. With Docker, your teams stay on the distros they already run, which means the adoption cost is measured in hours, not quarters.
One of our customers at Attentive (Stephen Commisso, Principal Engineer) captured their experience in the phrase “200 services – zero drama”, when describing their DHI rollout:
“The rollout was completely transparent to product teams. We had zero issues across over 200 services, which was particularly impressive since we were simultaneously switching Linux distributions from Ubuntu to Debian. All the heavy lifting happened during POC.”
We build every system package from source, for the distros you already use
With the launch of Hardened System Packages, Docker builds tens of thousands of Alpine and Debian system packages from source in a SLSA Build Level 3 pipeline with cryptographic signed, full provenance. This fundamentally changes the CVE equation.
Other vendors also claim to build system packages from source. The difference is that they build them for proprietary Linux distributions that have not had the benefit of independent community scrutiny and that customers have never run in production.
Docker builds packages for Alpine and Debian, the distributions your teams already operate, already test against, and already trust. Alpine and Debian are vast ecosystems that have independent maintainers, public mailing lists, coordinated disclosure with upstream projects, and volunteer security teams that operate independently of any commercial interest. You get the security benefit of from-source patching without the compatibility cost of adopting an unfamiliar OS.
We didn’t stop at near-zero CVEs, we made every image independently verifiable
Docker’s approach to container security is built on five pillars: minimal attack surface, verifiable SBOMs, secure build provenance, exploitability context, and cryptographic verification. We distilled our product development philosophy to these ideas, because we think your security posture depends on it. Not every vendor in the hardened image market shares this philosophy.
Most vendors in this space optimize for one metric: a clean CVE scan result.
Docker obsesses over near-zero CVEs too, but we went further: we built an attestation infrastructure that gives your security team, auditors, SOC, and change advisory boards machine-readable, cryptographically signed evidence for every question they will ask about an image.
We add 17 signed attestations to every single one of our 2000+ images in the DHI catalog, because that is what it takes to give you independent verifiability:
Question
DHI included attestation(s)
What this attestation is
Why it matters to you
What’s in this image?
CycloneDX SBOM, SPDX SBOM
Machine-readable inventory of every package, version, and transitive dependency. 
First thing auditors request during compliance reviews. Both formats are included so you don’t have to convert for different toolchains.
How was it built, and can I prove it?
SLSA provenance v1, SLSA verification summary, Scout provenance, DHI Image Sources
Cryptographic proof linking every image to its exact source definition. 
Required by supply chain security policies. Used by incident responders during forensics to verify whether an image was legitimately built or injected.
What vulnerabilities exist, and what’s been assessed?
CVEs v0.1, CVEs v0.2, VEX, Scout health score
CVE scan results and per-CVE exploitability justifications attached to the image itself. 
When your GRC team prepares a FedRAMP POA&M or your security team triages a new advisory, the evidence is already on the artifact.
Is it compliant?
FIPS compliance, STIG scan
FIPS evidence and OpenSCAP-generated STIG results
Ready artifacts for FedRAMP, PCI DSS, and HIPAA audits. Typically the most expensive artifacts to produce manually. Docker generates them automatically.
Has it been checked for non-CVE risks?
Secrets scan, Virus scan, Tests
Confirms no leaked credentials, no known malware, and that the image functions as expected. 
These are the checks SOC teams and security review boards require before approving production deployment. Docker runs them on every build.
What changed?
Changelog
Signed record of what was added, removed, or patched between versions.
Change advisory boards need this to approve updates. Without it, your team is diffing images manually.
Attestations answer questions about the image, the next set of questions are about the vendor.
What to ask your vendor, and what we found when we asked ourselves the same questions
In a fast-moving ecosystem, CVEs will occasionally get missed, advisories will have gaps, and no vendor operating at scale will have a flawless record. What matters is whether the gaps reveal isolated incidents or a pattern. The following questions are worth asking every vendor, including Docker.
What is the extent of your vendor’s commitment to patching?
Ask your vendor how far they go to resolve vulnerabilities. Docker continuously patches CVEs across both Debian, Alpine, and several major OSS software projects, rebuilding tens of thousands of system packages and several thousand images from source. That is a significant engineering and operational investment that most vendors avoid, because it is easier to build images for a single proprietary OS.
Docker’s commitment doesn’t end at images in our catalog. When a fix doesn’t exist upstream, there are many examples of Docker’s security team creating one. For CVE-2025-12735, a 9.8 CRITICAL RCE in Kibana’s dependency chain, the affected library was unmaintained and had no patch. Docker created the fix, shipped it to customers, and contributed it to LangChain.js. The fix was released as a public npm package on November 17, 2025.
One vendor we looked at has a published CVE policy of 7-day remediation for critical CVEs, once a qualifying patch is publicly available. In this instance, their fix appeared several weeks after that qualifying patch was created by Docker and shipped by the upstream project.
This level of upstream commitment is built into how our security team operates. Docker has been a MITRE CVE Numbering Authority since 2022, part of a sustained investment in teams’ ability to identify, disclose, and fix vulnerabilities at the source.
What assurances do you have about the completeness of your SBOMs?
Ask whether your vendor’s SBOM includes compiled dependencies (Rust crates, Go modules, JavaScript packages), or just system-level packages. Ask whether you can independently verify SBOM completeness against the project’s actual dependency manifest. Docker’s SBOMs include every compiled dependency. We’ve examined images from other vendors, and as one example for Vector (observability pipeline compiled from hundreds of Rust crate dependencies) one vendor’s SBOM did not appear to include those dependencies.
If a dependency isn’t in the SBOM, vulnerabilities in that dependency are invisible to the customer’s scanner and unverifiable by the customer’s security team. When Docker’s security team identified a High-severity CVE in Vector’s Rust dependencies, it was patched and shipped the same evening.
Does your vendor’s advisory feed surface every known CVE for the packages it ships?
Ask whether you can scan the vendor’s images with a third-party scanner against public advisory data, without relying on the vendor’s own advisory feed, and still get consistent results.
Docker recommends validating with Grype, Trivy, Wiz, or Mend. When we examined a vendor’s node image: CVE-2025-9308 and CVE-2025-8262 (both affecting yarn 1.22.22) were present in the shipped image but neither appeared on the vendor’s vulnerability page or in their security advisory feed. Docker’s hardened system package for yarn 1.22.22 is built from source with patches applied for both CVEs.
If your vendor’s advisory feed has gaps, your scanner inherits those gaps, and your security team is making decisions based on incomplete data.
When a CVE is assessed as unexploitable, does your vendor provide an auditable justification?
Not every CVE warrants a patch, and every vendor makes that judgment call. The question is whether your team can see the reasoning. Docker’s security team evaluates exploitability in the context of each minimal container image and publishes every assessment transparently.
Some vendors may set advisory version ranges to values real packages never match, thereby making CVEs invisible to scanners, and not providing a justification or an audit trail.
Docker uses VEX, the CISA-backed standard for communicating exploitability, which provides a per-CVE, machine-readable justification that every customer can read and audit.
We took on the parts of supply chain security others leave behind
Beyond multi-distro support, from-source patching, and transparency, we made a set of choices that compound into a distinctive, secure, simple experience for you.
Most vendor guarantees stop at the edge of the base image. Docker takes full ownership of your customized images: you add what your environment needs, and when a CVE is patched upstream, Docker automatically rebuilds your customized image and our SLA propagates to every artifact we produce. Your customizations don’t void the security guarantee. We’ve also opened up our hardened systems packages repo so you can use those hardened packages in your own bespoke containers. 
We will be extending this same rigor to language libraries next. The dependencies your application pulls in through npm, pip, or Maven will carry the same provenance and patching guarantees as the OS layer beneath them.
And for organizations running software that upstream has stopped supporting, Extended Lifecycle Support continues delivering security patches for up to five years past end-of-life, so teams can maintain their security posture while upgrading on their own timeline.
Come join the movement
A year ago, 500k daily pulls of the DHI catalog and a million builds running regularly felt like a milestone. Today, this is the baseline.
None of this would have happened without the teams who trusted us early and pushed us hard, including Adobe, Crypto.com, Attentive, and many others. Projects like n8n.io helped us understand what it takes to operate at scale. Partners like Socket.dev, Snyk, and Mend.io are building security workflows on top of this foundation.
We are continuing to listen, iterate, and do the hard things that are better for you, because that matters. If you are thinking about supply chain security, especially given the quantity and intensity of supply chain risks AI agents being to the mix, now is the time to raise your baseline with Docker.
Explore the Docker Hardened Images catalog and secure your supply chain here: https://www.docker.com/products/hardened-images/
For every team and developer, the open source DHI Community tier provides and immediately upgraded security posture. For businesses, we have a wide range of options that will work for your specific needs.
More resources:
DHI documentation: https://docs.docker.com/dhi/ Watch: Why n8n.io moved to DHI Read: Medplum’s step-by-step DHI adoption playbook

View the full article
Microsoft today pushed software updates to fix a staggering 167 security vulnerabilities in its Windows operating systems and related software, including a SharePoint Server zero-day and a publicly disclosed weakness in Windows Defender dubbed “BlueHammer.” Separately, Google Chrome fixed its fourth zero-day of 2026, and an emergency update for Adobe Reader nixes an actively exploited flaw that can lead to remote code execution.

Redmond warns that attackers are already targeting CVE-2026-32201, a vulnerability in Microsoft SharePoint Server that allows attackers to spoof trusted content or interfaces over a network.
Mike Walters, president and co-founder of Action1, said CVE-2026-32201 can be used to deceive employees, partners, or customers by presenting falsified information within trusted SharePoint environments.
“This CVE can enable phishing attacks, unauthorized data manipulation, or social engineering campaigns that lead to further compromise,” Walters said. “The presence of active exploitation significantly increases organizational risk.”
This flaw drops alongside a separate SQL Server remote code execution vulnerability (CVE-2026-33120), notes Ryan Braunstein, manager of Security and IT at Automox.
“One bug allows an attacker to get into your SQL instance from the network,” Braundstein said. “The other lets someone already inside promote themselves to full control.”
Microsoft also addressed BlueHammer (CVE-2026-33825), a privilege escalation bug in Windows Defender. According to BleepingComputer, the researcher who discovered the flaw published exploit code for it after notifying Microsoft and growing exasperated with their response. Will Dormann, senior principal vulnerability analyst at Tharros, says he confirmed that the public BlueHammer exploit code no longer works after installing today’s patches.
Satnam Narang, senior staff research engineer at Tenable, said April marks the second-biggest Patch Tuesday ever for Microsoft. Narang also said there are indications that a zero-day flaw Adobe patched in an emergency update on April 11 — CVE-2026-34621 — has seen active exploitation since at least November 2025.
Adam Barnett, lead software engineer at Rapid7, called the patch total from Microsoft today “a new record in that category” because it includes nearly 60 browser vulnerabilities. Barnett said it might be tempting to imagine that this sudden spike was tied to the buzz around the announcement a week ago today of Project Glasswing — a much-hyped but still unreleased new AI capability from Anthropic that is reportedly quite good at finding bugs in a vast array of software.
But he notes that Microsoft Edge is based on the Chromium engine, and the Chromium maintainers acknowledge a wide range of researchers for the vulnerabilities which Microsoft republished last Friday.
“A safe conclusion is that this increase in volume is driven by ever-expanding AI capabilities,” Barnett said. “We should expect to see further increases in vulnerability reporting volume as the impact of AI models extend further, both in terms of capability and availability.”
Finally, no matter what browser you use to surf the web, it’s important to completely close out and restart the browser periodically. This is really easy to put off (especially if you have a bajillion tabs open at any time) but it’s the only way to ensure that any available updates get installed. For example, a Google Chrome update released earlier this month fixed 21 security holes, including the high-severity zero-day flaw CVE-2026-5281.
For a clickable, per-patch breakdown, check out the SANS Internet Storm Center Patch Tuesday roundup. Running into problems applying any of these updates? Leave a note about it in the comments below and there’s a decent chance someone here will pipe in with a solution.
View the full article
Samsung raised the prices of several of its smartphones and tablets in the United States overnight, likely due to increasing costs caused by ongoing memory shortages.


As shared by PhoneArena, the 512GB Galaxy Z Flip increased by $80 from $1,219.99 to $1,299.99, while the S25 Edge and S25 FE went up $80 and $40, respectively. Samsung did not increase the cost of its current-generation S26 smartphones because those were priced higher than the prior-generation models from launch, but the company did also increase the cost of the 512GB and 1TB Galaxy Z Fold 7 earlier this month.

Samsung's tablet prices increased for the full U.S. lineup, including the latest Galaxy Tab S11 and the Galaxy Tab S11 Ultra models. The base models are $100 more expensive at $900 for the S11 and $1,299 for the S11 Ultra, while higher-end models went up even more. The 1TB S11 Ultra is now $1,899.99, which is a $280 increase.

Samsung didn't comment on the price increase, but the quiet price hike suggests that Samsung is going to need to charge more for upcoming devices that are coming out later this year.

As a major smartphone manufacturer, Samsung has not been able to weather rising costs without raising prices, and that could be a sign that Apple's upcoming devices could also be more expensive than they would have been without hardware shortages.

The MacBook Air and MacBook Pro models that came out earlier this year are more expensive than their predecessors, though Apple increased storage to justify the price hike.

Apple already removed the 512GB RAM upgrade for the Mac Studio, and started charging $400 more for the 256GB RAM upgrade. Apple also recently stopped accepting orders for some ‌Mac Studio‌ and Mac mini configurations with higher amounts of RAM. For machines still in stock, shipping times are extraordinarily high.

Hardware makers like Apple are dealing with high demand for memory and storage, which has been caused by demand from artificial intelligence data centers. Manufacturers are prioritizing AI chip production over chips designed for consumer products because large data center contracts are more profitable. Chipmakers like Samsung, TSMC, and SK Hynix are unable to keep up with demand even while operating at full capacity, and the lack of supply mixed with rising demand has led to price hikes.

During Apple's January earnings call, Tim Cook said that memory costs didn't impact the company's gross margin in the first fiscal quarter of the year, but would have a "bit more of an impact" during the second fiscal quarter. Apple is set to hold its Q2 earnings call on April 30.

Cook said that Apple is looking at a "range of options" to deal with rising prices over the long term if needed, and Apple is seeking supplier price cuts in other areas to offset the increase. Apple apparently agreed to pay Samsung twice as much for LPDDR5X memory chips for ongoing iPhone 17 production.

Apple analyst Ming-Chi Kuo said in January that he expects Apple to keep iPhone 18 starting prices flat despite having to pay more for components. He said Apple could absorb the costs to gain market share, and make up some of the money on its services side. Apple plans to launch a new foldable iPhone this year, and its rumored $2,000+ price tag could also pad Apple's margins.

Apple has been planning M5 updates for the ‌Mac Studio‌ and the ‌Mac mini‌, and it is unknown how the memory shortages and long shipping times for current machines will impact those plans.Tag: Samsung
This article, "Samsung's U.S. Price Increases Add to Concerns About Rising Apple Device Costs" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Security teams are stretched thin. Alerts never stop, attackers move faster, and expectations for uptime and resilience keep rising. For many IT and security leaders, Managed Detection and Response (MDR) has become less of a “nice to have” and more of a practical way to stay ahead.
But outsourcing MDR is not just about handing alerts to someone else. The real question is whether MDR helps you build cyber resilience, the ability to detect threats quickly, contain impact, and keep the business running.
Here are four questions to ask when deciding whether MDR belongs in your security strategy.
1. Do you have the coverage to detect threats 24/7?
Most attacks do not happen conveniently during business hours. They start late at night, on weekends, or during holidays when teams are understaffed or offline. If alerts sit unreviewed for hours, attackers gain time to escalate privileges, move laterally, and cause damage.
MDR closes this gap by providing continuous monitoring across endpoints, identities, and cloud environments. Instead of relying on best‑effort internal coverage, MDR ensures threats are reviewed and acted on around the clock.
This is a foundational part of cyber resilience. Faster detection means less dwell time, fewer affected systems, and easier recovery. Without 24/7 coverage, resilience becomes reactive rather than intentional.
2. Can your team separate real threats from noise?
Alert fatigue is one of the biggest barriers to effective security. Tools generate volumes of signals, but not all alerts represent real risk. When everything looks critical, teams either burn out or miss the alerts that matter most.
MDR helps by applying human expertise and threat intelligence to validate alerts, investigate behavior, and confirm whether activity is truly malicious. Instead of chasing every signal, your team receives clear guidance on what needs action and why.
Adlumin MDR™ supports this by correlating identity, endpoint, and network activity, then prioritizing threats based on real attacker behavior. The result is fewer distractions and faster, more confident response.
From a resilience perspective, this matters because a delayed or incorrect response often causes more disruption than the attack itself.
3. When an attack happens, can you contain it quickly?
Detection alone does not equal resilience. The difference between a security incident and a business‑level disruption often comes down to how quickly you can contain the threat.
Effective MDR does more than raise alerts. It helps security teams take action, isolating compromised systems, stopping malicious processes, and preventing spread before attackers reach critical assets.
For organizations without a full in-house SOC, MDR provides response capabilities that would otherwise require significant staffing investment. For MSPs, it enables consistent containment across many client environments without scaling headcount linearly.
When MDR is integrated with endpoint and identity controls, response becomes faster and more coordinated. This is a key step in minimizing attack impact and maintaining business continuity.
4. Does MDR fit into a broader cyber resilience strategy?
MDR is most effective when it is part of a before‑during‑after approach to cyber resilience.
Before an attack, reduce exposure with patching, configuration management, and least‑privilege access. Tools like N-central RMM™ help automate these fundamentals. During an attack, MDR detects and contains malicious activity in real time, reducing blast radius. After an attack, fast recovery determines whether operations resume quickly or stall. Cove Data Protection™ supports resilience with cloud‑first, immutable backups and rapid restore options. MDR plays a critical role in the “during” phase, but its value increases when it is connected to prevention and recovery. Resilience is not about any single control. It is about how well your controls work together under pressure.
Outsourcing MDR is about resilience, not just resources
The decision to outsource MDR is rarely about replacing your security team. It is about extending capabilities, improving response speed, and reducing the operational risk that comes from limited coverage and alert overload.
If your team struggles with 24/7 monitoring, alert validation, or rapid containment, MDR can be a practical way to strengthen resilience without adding complexity or headcount.
Cyber resilience depends on how quickly you can detect, respond, and recover. MDR helps close those gaps so attacks stay contained and the business keeps moving.
Check out the new 2026 State of the SOC Report and get insights based on real-world alerts from the Adlumin MDR SOC.
View the full article
The new N-able and Futurum Report reveals how AI is reshaping cyber resilience as it accelerates both business innovation and adversarial tradecraft. Attackers are scaling their operations with unprecedented speed, leveraging automation to bypass traditional defenses. For IT security leaders and MSP owners, the days of relying on static, perimeter-based security are over.
To stay ahead, we must look beyond simply keeping attackers out. The future belongs to those who prioritize continuous cyber resilience—the ability to withstand, adapt to, and recover from threats in real-time.
Here is where the industry is heading and how AI-powered platforms are redefining defense strategies for the next 3–5 years.
1. AI is both the weapon and the shield
We are entering an era of AI-versus-AI warfare. Threat actors are already using generative AI to craft sophisticated phishing campaigns, automate vulnerability scanning, and execute attacks at machine speed.
However, AI is also our most powerful ally. Modern defense isn’t about human analysts staring at screens; it’s about leveraging AI for rapid detection, automated triage, and predictive insights. By processing vast amounts of telemetry data, AI can identify subtle anomalies that human teams might miss, allowing for preemptive neutralization of threats before they cause damage.
See how AI is altering the landscape in the new N-able and Futurum report, Cybersecurity in the Age of AI: Moving from Fragile to Resilient. Get key insights on building a modern framework for business resilience.
2. From perimeter security to continuous cyber resilience
The “castle and moat” approach is obsolete. In a world of hybrid work and distributed cloud environments, the perimeter has dissolved. The new standard is continuous cyber resilience.
This shift moves the goalposts from “prevention” to “continuity.” It focuses on always-on monitoring, continuous validation of security postures, and real-time response capabilities. The objective is to ensure that even when a breach attempts to occur, the business keeps running. This requires a mindset shift: resilience is not a product you buy, but a state of operation you maintain.
3. Platform-based security is the new norm
Managing dozens of disparate point solutions is a logistical nightmare that introduces security gaps. The industry is rapidly consolidating toward integrated platforms that offer a unified view of the threat landscape.
For MSPs and enterprise IT leaders, platform-based security offers streamlined operations and better data correlation. By unifying endpoint management, threat detection, and backup solutions, organizations can reduce complexity and improve their mean time to response (MTTR).
4. Human-centric security operations
Despite the rise of AI, the human element remains critical—but the role is changing. We are moving toward human-centric security operations where AI handles the low-value, repetitive work, freeing up analysts to focus on high-impact strategic decisions.
Automation is the key to alleviating burnout. By automating alert triage and routine remediation tasks, we can improve the analyst experience and ensure that human expertise is applied where it matters most: complex threat hunting and strategic resilience planning.
Explore the benefits of integrating AI into your cybersecurity strategy and learn how it revolutionizes your security operations across crucial threat detection and response stages.
5. Regulation, reporting, and resilience mandates
Regulatory pressure is mounting. Governments, insurers, and customers are no longer satisfied with improved security promises; they demand demonstrable resilience.
Expect to see stricter mandates around incident reporting timelines and resilience metrics. Organizations will need to prove not just that they are secure, but that they can recover quickly and effectively from disruptions. This transparency will become a competitive differentiator in the market.
Compliance is not just about checking boxes. Our solutions, like Adlumin MDR, offer enhanced compliance support with one-touch reporting.
Preparing your organization for the future
The next few years will separate the reactive from the resilient. To future-proof your defense, consider these actionable steps:
Assess platform readiness: Evaluate whether your current toolset supports integration and automation or if it creates silos. Invest in automation: Look for opportunities to automate routine tasks and free up your team for strategic work. Build resilience metrics: Move beyond simple uptime stats. Measure your ability to recover and adapt to new threats. Align teams: Ensure that IT and security teams are working in lockstep, supported by a unified platform vision. The N-able vision for AI-powered resilience
At N-able, we are committed to fueling your IT success by staying ahead of the curve. We see a future where our platform evolves continuously, leveraging AI not just to detect threats, but to anticipate them.
Our vision is to provide partners and customers with an ecosystem that supports continuous threat hunting, intelligent automation, and seamless recovery. As the threat landscape evolves, so do we—ensuring you have the foundation needed to secure your clients effectively.
The right partner makes all the difference. To learn more, visit us here.
View the full article
Apple has two new Apple Watch activity challenges coming up, celebrating Earth Day and International Dance Day. The Earth Day activity challenge will launch on Wednesday, April 22, while the Dance Day challenge will take place a week later on Wednesday, April 29.


To complete the Earth Day challenge, Apple Watch owners will need to complete a workout that lasts for 30 minutes or longer.

The International Dance Day award can be earned by completing a Dance workout of 20 minutes or more.

Apple Watch owners who earn the Earth Day award will unlock a badge in the Fitness app, along with animated stickers.








The Dance Day challenge will also unlock a special Fitness app badge and accompanying stickers.








Earth Day and International Dance Day are two events that are always part of Apple's ongoing Apple Watch activity challenge schedule. They follow the February Heart Month activity challenge and the January New Year challenge.Related Roundup: Apple Watch 11Tag: Activity ChallengeBuyer's Guide: Apple Watch (Neutral)
This article, "Apple Watch Earth Day and International Dance Day Activity Challenges Launching Later This Month" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
A fake Mac app designed to look like the real thing snuck past Apple's app review team, costing users $9.5 million in cryptocurrency.


According to CoinDesk, a fake macOS version of the Ledger Live crypto wallet app scammed people into handing over access to their cryptocurrency wallets. More than 50 people fell victim to the fake app between April 7 and April 13.

Ledger has an official Mac app, but it is distributed via the Ledger website and not through the Mac App Store. The real app does not ask users to enter their seed phrases like the fake app did, nor do other legitimate cryptocurrency apps. The stolen money was routed through the KuCoin crypto exchange, and hackers used a mixing service known as AudiA6, which charges high fees to launder cryptocurrency.

Three of the victims lost seven-figure sums, which is an unusually high amount of money to lose in a fake app scam. ZachXBT, who investigated the scam and shared the info on Telegram, suggested Apple could be subject to a class-action lawsuit in the future due to the amount of money lost.

Apple removed the fake Ledger Live app from the ‌Mac App Store‌, but it was live for approximately two weeks. It is not known how it passed Apple's app review process, and Apple hasn't commented.Tags: Cryptocurrency, Mac App Store
This article, "Apple Removes Fake Crypto Wallet App That Stole $9.5 Million From Mac Users" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Apple today launched its new all-in-one Apple Business platform, debuting the refreshed Apple Business web portal and accompanying app.


Apple Business aggregates several of Apple's prior business-focused products, like Apple Business Essentials, Apple Business Manager, and Apple Business Connect. The service offers organizations a unified platform for managing devices, employees, communications, and customer engagement across the Apple ecosystem.

Companies can take advantage of built-in mobile device management tools (MDM) for configuring device settings, security policies, available apps, and user groups from one location. With a simplified "Blueprints" option, employers can preconfigure devices purchased from Apple or authorized retailers with settings and apps for zero-touch deployment. Employees can use Apple Business to install work-related apps, request support from employers, and contact colleagues through a company directory.

Managed Apple Accounts provide "cryptographic separation" between personal and work data, so employees don't need to deal with multiple devices. Provisioning can be automated with providers like Google Workspace and Microsoft Entra ID.

Apple Business includes integrated email, calendar, and directory services linked to custom domains, plus customer engagement tools. Businesses can manage how their brand and locations show up across Apple services using brand profiles in Safari, Siri, and Spotlight, branded communication in Apple Mail, order tracking in Wallet, customizable place cards in Apple Maps, Tap to Pay branding, and more.

Apple is gearing up to introduce ads in ‌Apple Maps‌ this summer, and ads can be purchased through Apple Business. Businesses can create ads that show up at the top of search results in Maps and in a Suggested Places feature coming in iOS 26.5. Ads will be limited to the U.S. and Canada at launch.

Apple Business is available in over 200 countries and regions. It is a free service for new and existing users, with optional paid add-ons for more iCloud storage and AppleCare+ and no monthly fees for device management. Apple is discontinuing Apple Business Essentials, Apple Business Manager, and Apple Business Connect now that its unified Apple Business platform is available.

The Apple Business companion app and email, calendar, and directory features require iOS 26, iPadOS 26, or macOS 26.Tag: Apple Business
This article, "Apple Launches New All-in-One Apple Business Platform for Device Management, Email, and Customer Engagement" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Apple today provided public beta testers with the second releases of upcoming iOS 26.5, iPadOS 26.5, macOS Tahoe 26.5, watchOS 26.5, and tvOS 26.5 updates for testing purposes. The public betas come a day after Apple provided the betas to developers.


After signing up for beta testing on Apple's beta site, public beta testers can download the updates using the Software Update section of the Settings app on each device.

iOS 26.5, iPadOS 26.5, and ‌macOS Tahoe‌ 26.5 include a new Suggested Places feature for recommending nearby locations to visit, and Apple is also gearing up to start showing ads in Maps.

Apple is testing end-to-end encryption for RCS messages between iPhone and Android users again, and there are proximity pairing, notification forwarding, and Live Activities for third-party wearables in the EU.Related Roundups: iOS 26, iPadOS 26, macOS TahoeRelated Forums: iOS 26, macOS Tahoe
This article, "Apple Releases Second iOS 26.5, iPadOS 26.5 and macOS Tahoe 26.5 Public Betas" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
European regulators have largely been frozen out of early access to Anthropic’s new Mythos model, Politico reports. The AI technology, aimed at cybersecurity use cases, is said to be able to identify and exploit technical vulnerabilities at a level that surpasses most humans — signaling a structural shift for CISOs and the cybersecurity industry.
For security reasons, Anthropic has chosen to initially limit access to a few selected players as part of Project Glasswing. These include primarily US technology giants such as Apple, Microsoft, and Amazon, who have been given the opportunity to analyze the model and address potential security flaws.
At the same time, European authorities have been largely left out. In the UK, the country’s AI Security Institute has been allowed to test the model and has already acted on its results. Within the EU, only Germany is currently reported to have initiated a dialogue with Anthropic, but without yet gaining access to the technology.
As a result, the EU’s influence over the model is limited, not least because it is not yet widely available. Several experts interviewed by Politico believe that this development is problematic, as private companies in practice decide how and when such powerful technology is shared, rather than independent authorities.
Germany’s chief cybersecurity official Claudia Plattner underscored to Politico the “pressing” question of whether a tool like Mythos “of such extraordinary power” will be available on the open market.
“That question, in turn, has profound implications for national and European security and sovereignty,” Plattner said.
View the full article
Two high-severity security vulnerabilities have been disclosed in Composer, a package manager for PHP, that, if successfully exploited, could result in arbitrary command execution. The vulnerabilities have been described as command injection flaws affecting the Perforce VCS (version control software) driver. Details of the two flaws are below - CVE-2026-40176 (CVSSView the full article
Blackmagic Design has announced a major update to its professional video editing and color correction software, DaVinci Resolve, including a new Photo page that aims to streamline image reframing and cropping.


DaVinci Resolve 21 extends the application's color grading toolset to still photography for the first time, meaning photographers can now apply primary color correction, curves, qualifiers, power windows, and node-based edits to stills, with changes held at the original source resolution. An additional LightBox view displays whole albums with grades applied, and Sony or Canon cameras can be tethered for direct capture into albums.

Unsurprisingly perhaps, much of this update centers on AI. A tool called IntelliSearch indexes media so editors can search for objects, spoken keywords, or specific faces. Meanwhile, CineFocus lets users shift a shot's focal point after recording and add bokeh, while a set of facial tools can age or de-age subjects, reshape features, and remove blemishes.

Two further additions, UltraSharpen and Motion Deblur, are aimed at salvaging soft or blurry footage.

Elsewhere in the app, keyframing gains four-point Bezier easing and the ability to adjust multiple clips at once, and Fusion effects can now be tweaked directly from the Cut and Edit pages. Text handling also picks up multi-language spell check, a font browser, emoji support, and character-level styling. The Cut page now has smart bins, while a new MultiMaster trim manager lets colorists generate multiple HDR and SDR deliverables from a single timeline.

Resolve 21 also introduces native support for OGraf HTML graphics and Lottie animations, so users can now drag .json and .lottie files directly into the media pool, where they will be treated like fully rendered animation clips. There's also a Picture in Picture effect, and expanded IntelliScript support for Final Draft and plain text screenplays. See the press release for further details on all the improvements and changes.

DaVinci Resolve 21 public beta is available now to download for free from the Blackmagic Design website, but we're still waiting for a general release date to be confirmed.Tag: Blackmagic
This article, "DaVinci Resolve 21 Adds Photo Editing and AI Search Tools" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Expercom has introduced new record low prices across the entire 2026 Studio Display XDR lineup this week, with up to $200 off these monitors. Expercom is a reliable Apple partner that sells both brand new and pre-owned Apple products at discounted prices.

Note: MacRumors is an affiliate partner with Expercom. When you click a link and make a purchase, we may receive a small payment, which helps us keep the site running.

In regards to the 2026 Studio Display XDR, these are all new models with deals that beat the current Amazon discounts by over $100 in some cases. Prices start at $2,749.00 for the Studio Display XDR with Standard Glass and VESA Mount Adapter, down from $2,899.00.

$150 OFFStudio Display XDR (Standard/VESA) for $2,749.00
$150 OFFStudio Display XDR (Nano-Texture/VESA) for $3,049.00
$200 OFFStudio Display XDR (Standard/Tilt and Height) for $3,099.00
$200 OFFStudio Display XDR (Nano-Texture/Tilt and Height) for $3,399.00

All models are in stock and ready to ship with the exception of the Nano-Texture Glass Studio Display XDR with tilt- and height-adjustable stand, which has a slight 7-14 day shipping delay. If you're shopping for the regular Apple Studio Display, Amazon currently has a few models at best-ever prices.

Head to our full Deals Roundup to get caught up with all of the latest deals and discounts that we've been tracking over the past week.



Deals Newsletter

Interested in hearing more about the best deals you can find in 2026? Sign up for our Deals Newsletter and we'll keep you updated so you don't miss the biggest deals of the season!




Related Roundup: Apple Deals
This article, "Apple's 2026 Studio Display XDR Drops to New Record Low Prices on Expercom" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
A joint status report filed yesterday in Apple's trade secrets lawsuit against YouTuber Jon Prosser and Michael Ramacciotti shows Prosser is still failing to comply with discovery, prompting Apple to seek a court order to compel him.


The latest filing, submitted to the U.S. District Court for the Northern District of California yesterday, covers developments since the parties' last update in February 2026. It notes that Apple served Prosser with document and deposition subpoenas on February 3, and that while he has provided some responsive materials, he has failed to fully respond to certain requests and has not responded at all to others. Apple has extended his deadline multiple times and says it has still not received the limited discovery it needs to understand the full scope of what confidential information Prosser and Ramacciotti obtained and how they got it. Apple says it now intends to file a Motion for an Order to Show Cause in the Northern District of Ohio to force his compliance.

The filing also reveals that Prosser has indicated he is retaining counsel and intends to move to set aside the default judgment entered against him in October 2025, after he missed the court deadline to respond to Apple's complaint. At the time, Prosser told The Verge he had "been in active communications with Apple since the beginning stages of this case," a claim Apple subsequently disputed in court documents.

Apple filed the lawsuit in July 2025, accusing Prosser and Ramacciotti of misappropriating trade secrets by gaining unauthorized access to a development iPhone belonging to former Apple software engineer Ethan Lipnik. According to Apple's complaint, Ramacciotti accessed the device while Lipnik was away and showed Prosser the contents over FaceTime, revealing details about what was then called iOS 19 and later unveiled at WWDC 2025 as iOS 26. Prosser published videos on his YouTube channel showing recreated renderings of the software's Liquid Glass design months before Apple's announcement. Lipnik was terminated for failing to follow Apple's policies for securing development devices.

Ramacciotti's posture in the case stands in contrast to Prosser's. According to the filing, he has allowed Apple to forensically review an additional device, agreed to supplement his interrogatory responses, and offered to sit for a follow-up deposition once Apple completes its third-party discovery, including any deposition of Prosser. Apple and Ramacciotti have been informally discussing a potential settlement since at least October. Apple is seeking monetary damages and an injunction barring both defendants from further disclosing any of the company's confidential information.

The parties have scheduled a further status update with the court for June 10, 2026.Related Roundups: iOS 26, iPadOS 26Tags: Jon Prosser, Apple LawsuitsRelated Forum: iOS 26
This article, "Jon Prosser Still Not Fully Cooperating in Apple's iOS 26 Trade Secrets Lawsuit" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Google has announced the integration of a Rust-based Domain Name System (DNS) parser into the modem firmware as part of its ongoing efforts to beef up the security of Pixel devices and push memory-safe code at a more foundational level. "The new Rust-based DNS parser significantly reduces our security risk by mitigating an entire class of vulnerabilities in a risky area, while also layingView the full article
Amazon this week has multiple discounts on the brand new M4 iPad Air, providing up to $100 off these brand new models.

Note: MacRumors is an affiliate partner with Amazon. When you click a link and make a purchase, we may receive a small payment, which helps us keep the site running.

Specifically, Amazon has up to $83 off the 11-inch M4 iPad Air and up to $100 off the 13-inch M4 iPad Air. All of these discounts have been automatically applied and do not require a coupon code or a Prime membership.

$48 OFF11-inch M4 iPad Air for $551.57
$50 OFF13-inch M4 iPad Air for $749.00

The new iPad Air features the M4 chip, C1X modem, and N1 networking chip, which brings support for Wi-Fi 7 and Bluetooth 6. In terms of design, the 2026 models are identical to the 2025 iPad Air tablets, with an edge-to-edge display, slim bezels, and aluminum chassis.

11-inch M4 iPad Air

128GB Wi-Fi - $551.57 ($48 off)
256GB Wi-Fi - $646.50 ($53 off)
512GB Wi-Fi - $836.50 ($63 off)
1TB Wi-Fi - $1,016.50 ($83 off)

13-inch M4 iPad Air

128GB Wi-Fi - $749.00 ($50 off)
256GB Wi-Fi - $836.50 ($63 off)
512GB Wi-Fi - $1,019.00 ($80 off)
1TB Wi-Fi - $1,199.00 ($100 off)

If you're on the hunt for more discounts, be sure to visit our Apple Deals roundup where we recap the best Apple-related bargains of the past week.



Deals Newsletter

Interested in hearing more about the best deals you can find in 2026? Sign up for our Deals Newsletter and we'll keep you updated so you don't miss the biggest deals of the season!




Related Roundup: Apple Deals
This article, "Apple's M4 iPad Air Available for Up to $100 Off on Amazon" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Chrome has been updated today with a Skills library that's designed to let Chrome users turn AI tasks into repeatable skills that can be used on any website.


Useful prompts you create for Gemini in Chrome can be saved as a Skill that can be accessed later with a single click. If you're shopping for skincare and ask Gemini about the ingredients in a product, for example, you can save the question as a Skill and then use it again later without needing to re-type the prompt.


Google provided the following examples of how testers have used the feature across different categories.

Skills can be saved directly from the chat history in Chrome (located in the side panel when Gemini is enabled), and recalled by typing a forward slash and the Skill name or clicking on the plus sign. The selected Skill will run on the page that's being viewed, along with other selected tabs.

Google is debuting the feature with a library of pre-written Skills for common tasks and workflows like viewing ingredients, finding a gift for someone, or making substitutions in a recipe. Pre-prepared Skills can be customized as needed.

When using a Skills prompt, Gemini will confirm before taking actions like adding an event to the calendar or sending an email, similar to other Gemini actions in Chrome. Skills are rolling out for Gemini in the desktop version of Chrome when the browser's language is set to U.S. English.Tags: Google, Google Chrome
This article, "Gemini in Google Chrome Gets a Skills Library for Saving Custom AI Prompts" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
A building permit filed in Zurich confirms Apple is planning a new retail store at Lintheschergasse 7, near the city's famous Bahnhofstrasse shopping street, with construction set to run through early 2027.


The permit, submitted to the city of Zurich in February and reviewed by Swiss Apple publication macprime, explicitly describes the project as a "Bahnhofstrasse Relocation" and includes Apple's internal designation "R159" for its Zurich store. The planned location sits at the corner of Lintheschergasse and Usteristrasse, adjacent to the Globus department store.

According to a person with knowledge of Apple's retail planning, the company's current Rennweg store was never intended as a permanent home. "The Rennweg store was planned as a temporary location, since the entrance has stairs and an elevator," the person told MacRumors. They said it was already an open secret when Apple left Bahnhofstrasse that the company was seeking a significantly larger space, and that a dedicated team was tasked with identifying possible buildings before the construction and planning department took over.

According to the permit documents, the ground floor, labeled "Sales," spans around 454 square meters, comparable to Apple's current Rennweg store. A first-floor space labeled "Backstage," likely office use, adds a further 521 square meters. Whether that upper floor will ultimately serve as retail space remains unclear. If both floors are used for sales, the total retail footprint would rank the store among the largest Apple locations in Europe, with Tagesanzeiger reporting an overall area of around 2,000 square meters. Estimated rent is around 1,500 Swiss francs per square meter per month.

The building at Lintheschergasse 7 is a listed municipal heritage structure, which limits how extensively the exterior can be altered. The permit describes a facade renovation covering the ground and first floors, including new metal panels in "Aluminium Champagne" between the shop windows and updated window frames, while the concrete pillar cladding matching the upper floors will remain unchanged. The works also involve a slight expansion of the ground-floor entrance area, eliminating a recessed corner entry in favor of a flush facade, adding 11 square meters and bringing the total building footprint to 965 square meters. A basement level will likely provide storage.

The construction timeline runs from early November 2026 to early May 2027, suggesting an opening sometime in summer 2027. Apple opened its first Zurich store on Bahnhofstrasse in May 2009, then relocated to Rennweg 43 in 2019. The planned Lintheschergasse location would mark the company's third Zurich address.

Separately, Apple is also planning a new Geneva store, with building permits there showing a dramatic glass facade design. Apple currently operates four retail stores in Switzerland, two in Zurich and one each in Basel and Geneva.Tags: Europe, Retail, Switzerland
This article, "Apple Plans One of Europe's Biggest Stores in Zurich" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Amazon and Globalstar have announced a definitive merger agreement under which Amazon will acquire the satellite operator. News of the deal also puts to bed questions about the fate of Apple's exclusive satellite connectivity partner, and how Apple fits into the bigger picture.


Alongside the acquisition, Amazon and Apple have also signed a separate agreement for Amazon's Leo satellite network to power current and future iPhone and Apple Watch satellite features, including Emergency SOS, Messages via satellite, Find My, and Roadside Assistance via satellite.

Amazon said it will continue supporting iPhone and Apple Watch models that use Globalstar's existing and upcoming low Earth orbit constellation, which is being built by MDA Space. Amazon also said it will work with Apple on future satellite services running on the expanded Leo network.
Amazon's acquisition deal is expected to close in 2027, subject to the usual regulatory approvals. Bloomberg had reported in October that Globalstar was exploring a sale and had held early talks with SpaceX before the Amazon discussions emerged. As we reported earlier this month, Apple's 20 percent stake in Globalstar was said to be a sticking point in Amazon's bid to acquire the company.

Apple is working on a series of new satellite connectivity features for the iPhone which will apparently require upgrades to Globalstar's infrastructure. They include Apple Maps via satellite, photos in Messages via satellite, connectivity in indoors environments, satellite over 5G, and a satellite API for third-party apps.Tags: Amazon, iPhone Satellite Features
This article, "Apple and Amazon Ink Satellite Deal Amid Globalstar Takeover" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Apple's iPhone development roadmap runs several years into the future and the company is continually working with suppliers on several successive iPhone models at the same time, which is why we often get rumored features months ahead of launch. The iPhone 18 series is no different, and we already have a good idea of what to expect for the iPhone 18 Pro and iPhone 18 Pro Max.


One thing worth noting is that Apple is reportedly planning a major change to its iPhone release cycle this year, adopting a two-phase rollout starting with the iPhone 18 series. That means the iPhone 18 Pro, iPhone 18 Pro Max, and iPhone Fold will be released in September 2026, followed by the iPhone 18 and iPhone 18e in spring 2027.


Overall Design

iPhone 17 Pro Style

Rumors suggest the iPhone 18 Pro lineup will largely retain the same design as the iPhone 17 Pro models. The rear camera system will look identical to the current generation, featuring a raised "plateau" with three lenses arranged in a triangle. Display sizes are also expected to remain unchanged, with the iPhone 18 Pro and iPhone 18 Pro Max continuing to use 6.3-inch and 6.9-inch panels, respectively – the same dimensions introduced with the iPhone 16 Pro series. iPhone 18 Pro models could drop the current two-tone look of the rear casing found on the iPhone 17 Pro in favor of a more seamless aesthetic. For the next-generation models, Apple has apparently updated the back-glass "replacement process" to minimize the color difference between the Ceramic Shield 2 glass and the aluminum frame, resulting in a more unified appearance.

Next-Level Battery Life

Thicker Chassis

The iPhone 18 Pro Max will feature a bigger battery for continued best-in-class battery life, claims a Chinese leaker. The Weibo user known as "Digital Chat Station" said that the ‌iPhone 18‌ Pro Max will have a battery capacity of 5,100 to 5,200 mAh. (The iPhone 17 Pro Max has the biggest ‌iPhone‌ battery to date at 5,088 mAh. Apple says it has a battery life of up to 39 hours.) According to another rumor, the body of the iPhone 18 Pro Max will be slightly thicker than the iPhone 17 Pro Max, raising the device's weight to around 243 grams. That would make the iPhone 18 Pro Max approximately 3 grams more than the iPhone 14 Pro Max, which is currently the heaviest model Apple has produced. A larger battery is the most likely cause.

Smaller Dynamic Island

Under-Screen Face ID?

Rumors continue to circulate about whether the iPhone 18 Pro models will introduce under-display Face ID, but reports remain divided on when the technology will actually arrive. The feature would move the TrueDepth camera system beneath the display, eliminating the need for the current Dynamic Island cutout.

According to Wayne Ma of The Information, Apple is targeting a design without a Dynamic Island, replacing it with a single pinhole camera in the upper-left corner of the screen. However, other sources dispute that claim. Display analyst Ross Young believes under-display Face ID is possible for the iPhone 18 Pro, but says a smaller Dynamic Island will still be present. Bloomberg's Mark Gurman has echoed this view, reporting that the new models will feature a slimmed-down Dynamic Island rather than removing it entirely. Apple is also said to be testing new camera miniaturization technology to reduce the size of the front-facing camera currently located within the Dynamic Island.

The Weibo leaker "Ice Universe" has claimed the Dynamic Island cutout on the iPhone 18 Pro models will be approximately 35% narrower than it is on the iPhone 17 Pro models. Specifically, they said it will have a width of around 13.5mm, down from around 20.7mm

Meanwhile, Chinese leaker Instant Digital has offered yet another version of events, saying the Dynamic Island will shrink in size, but that under-display Face ID and camera technology won't debut this year. The latest word on the subject is that Apple is weighing two options for the iPhone 18 Pro's Dynamic Island, and a final decision has yet to be made. One option apparently retains the existing screen mold from the iPhone 17 Pro, while the other introduces a significantly smaller "Mini ‌Dynamic Island‌" enabled by moving the Face ID receiver and transmitter components beneath the display.

A20 Pro Chip

2nm Process

The iPhone 18 Pro models will use Apple's A20 chip, based on TSMC's 2nm process for power and efficiency improvements. A move to 2nm fabrication increases transistor density, which will enable higher performance. The A20 series is expected to deliver roughly a 15 percent speed gain and about 30 percent better efficiency compared with the A19 series used in Apple's iPhone 17 models.

Apple's A20 chip will be packaged with TSMC's Wafer-Level Multi-Chip Module (WMCM) technology, suggesting at least some A20 chips will have RAM integrated directly onto the same wafer as the CPU, GPU, and Neural Engine, rather than sitting adjacent to the chip and connected via a silicon interposer. This could contribute to faster performance for both overall tasks and Apple Intelligence, and longer battery life from improved power efficiency.

C2 Modem

Replacing Qualcomm

Apple plans to include its next-generation C2 modem in the iPhone 18 Pro models, according to supply chain analyst Jeff Pu. The chip will succeed the C1 modem, which debuted in the lower-cost iPhone 16e as Apple's first in-house cellular modem, and the C1X modem chip in the iPhone Air, which Apple says is up to 2× faster than the C1. The C2 is expected to bring faster speeds, improved power efficiency, and support for mmWave 5G in the United States – a feature missing from the C1 and C1X.

Apple's modem roadmap is part of a long-term strategy to reduce reliance on Qualcomm, which currently supplies 5G modems for the rest of the iPhone lineup. The company has been working on developing its own cellular chips for years, aiming for deeper integration and greater control over power management and performance.

New Camera Sensor

Samsung-Made

Samsung is working on a new three-layer stacked image sensor, reportedly intended for the iPhone 18. The sensor, referred to as PD-TR-Logic, integrates three layers of circuitry, which would improve camera responsiveness, reduce noise, and increase dynamic range. The leak comes from a source known as "Jukanlosreve," who claims the sensor is being developed specifically for Apple's 2026 iPhone lineup. Sony has long been Apple's sole image sensor supplier, so Samsung's entry would be a big shift in the iPhone's camera supply chain.

Variable Aperture

DSLR-Style

Apple intends to equip this year's iPhone 18 Pro models with a variable aperture lens, according to reports. Weibo-based leaker Digital Chat Station claims the main rear camera – what Apple calls the 48-megapixel Fusion camera – on both iPhone 18 Pro models will offer variable aperture, which would be a first for the iPhone. A variable-aperture system physically adjusts the lens opening, letting more light in for low-light shots or narrowing the opening for brighter scenes and deeper depth of field.

The main cameras on the iPhone 15 Pro, 16 Pro, and 17 Pro all use a fixed ƒ/1.78 aperture, where the lens is permanently set to its widest setting. With a variable lens, the iPhone 18 Pro would allow users to manually shift the aperture, similar to on a DSLR camera. This would mean more control over depth of field, enabling sharper focus on subjects or smoother background blur. Industry analyst Ming-Chi Kuo said in November 2024 that Apple's iPhone 18 Pro models will get the feature.

5G Satellite Internet

Non-Terrestrial Data

According to a report by The Information, Apple plans to add support for 5G networks that operate via satellites rather than Earth-based towers as early as next year. This advancement would allow future iPhones to gain full internet connectivity through satellite, not just limited emergency features.

If Apple meets the 2026 target, the first devices to feature 5G satellite internet would likely be the iPhone 18 Pro, iPhone 18 Pro Max, and the long-rumored foldable iPhone. Apple partners with Globalstar for its iPhone satellite features, but there is currently no service that delivers full 5G satellite internet directly to a smartphone, and the report did not specify who would supply it.

Simplified Camera Control

New Design

Apple is reportedly working to simplify the Camera Control button's design on iPhone 18 models in order to reduce costs. The current Camera Control button on iPhone 17 models uses both capacitive and pressure sensors beneath a sapphire crystal surface. The capacitive layer detects touch gestures, while the force sensor recognizes different pressure levels for taps, presses, and swipes.

However, according to the Weibo-based account Instant Digital, Apple will remove the capacitive sensing layer and retain only pressure sensing recognition in the second iteration to achieve all Camera Control functions on the iPhone 18. The simplified version is not about reducing functionality in the button, but about saving money. The current solution is said to be very expensive for Apple and is generating costly after-sales repairs.

New Colors

Three in Testing

In February 2026, Bloomberg's Mark Gurman reported that Apple is testing a deep red finish for the iPhone 18 Pro and iPhone 18 Pro Max. Rumors of purple and brown finishes have also circulated, but Gurman believes those are just variants of the same red idea. The iPhone 14 Pro and iPhone 14 Pro Max were previously available in Deep Purple, and Apple has never released an iPhone in a genuinely brown color. According to a Chinese leaker, Apple's iPhone 18 Pro models won't come in black this year. If the rumor is true, it will be the second consecutive year Apple has ditched what was arguably its most classic color option for the Pro lineup.Related Roundup: iPhone 18 Pro
This article, "10 Reasons to Wait for the iPhone 18 Pro" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
The difference between a traditional SOC and an agentic SOC is not just technology. It is a fundamental redesign of how security work gets done. In this blog, we examine how AI agents change triage, investigations, escalation, and response, and why many organizations are moving toward agentic SOC models in a phased, controlled way.
View the full article
Blackmagic has announced a new version of its URSA Cine Immersive camera, the first commercial camera system designed to capture 3D content for the Vision Pro.


The URSA Cine Immersive 100G adds 100G Ethernet to the original camera to deliver the bandwidth needed to output live immersive video for the first time.

Blackmagic Design also announced the Blackmagic URSA Cine Live Encoder, a live processor module ($1,645) that compresses live immersive video into Apple ProRes for output as SMPTE-2110-22 IP video, allowing users to combine the stereo, high frame rate image streams into a single 100G Ethernet connection.

However, the capability is costly – Blackmagic is asking $29,145 for the device, which will be available in Q3 2026. The original Blackmagic URSA Cine Immersive remains available on the Blackmagic website for $27,495, down from the $30,000 price tag it carried when it was first unveiled in 2024.

Both cameras have a custom stereoscopic 3D lens system with dual 8K sensors, and can capture a 180-degree field of view with spatial audio support at up to 90 frames per second. Captured content features an 8,160 x 7,200 resolution per eye, and there are 16 stops of dynamic range for detail and color accuracy in every frame.

Dual 5-inch HDR touchscreens are also included, along with an external color status LCD screen. There are several other connectivity options aside from Ethernet, including 12G-SDI out, USB-C, and XLR audio ports, plus an 8-pin Lemo connector for power.


Blackmagic says URSA Cine Immersive has been used on a number of high-profile immersive productions, including MotoGP: Tour De Force, Debut at the BBC Proms, an upcoming documentary featuring Real Madrid, and NASA's recent Artemis II launch.Related Roundup: Apple Vision ProTag: BlackmagicBuyer's Guide: Vision Pro (Buy Now)Related Forum: Apple Vision Pro
This article, "Blackmagic Debuts $29K+ URSA Cine Immersive 100G for Vision Pro" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Artificial intelligence is rapidly reshaping how security teams detect and hunt cyber threats by helping analyze vast volumes of security data, uncovering subtle signs of malicious activity, and identifying potential attacks faster than traditional tools or human analysts alone.
Analyst firm Gartner expects that by 2028, 50% of threat detection, investigation, and response (TDIR) platforms — including technologies such as EDR, XDR, SIEM, and SOAR — will incorporate agentic AI capabilities, up from less than 10% in 2024. The firm says AI could help organizations strengthen threat detection, incident response, and containment while also helping security teams bridge persistent skills shortages and reduce reliance on scarce cybersecurity talent.
A matter of scale
Much of AI’s impact in threat detection is tied to its ability to process telemetry at a scale that human teams would find challenging, if not impossible, to manage, according to security experts.
Modern IT environments can generate billions of logs and events each day across endpoints, networks, cloud services, and identity systems. Machine learning models can correlate those signals in near real-time, and identify behavioral anomalies — such as unusual login patterns, suspicious lateral movement, or data exfiltration attempts — that might otherwise remain buried in the noise.
Many enterprise security teams expect such capabilities to significantly bolster their detection capabilities. In a 2025 survey that Anvilogic conducted in collaboration with the SANS Institute, 45% of respondents said their organizations have already integrated AI into their threat detection workflows; 88% believed AI would play a major role in detection engineering within the next three years.
Organizations are already using AI to automate many of the routine tasks traditionally handled by Tier 1 and Tier 2 analysts, says Martin Sordilla, senior technology and security architect at Accenture. Much of this work involves reviewing logs, triaging alerts, identifying indicators of compromise, correlating events, and reaching out to system owners during investigations. AI can significantly accelerate these processes — automating tasks such as alert triage, documentation, evidence collection, and chain-of-custody tracking, he adds.
Organizations are already seeing efficiency gains of roughly 40-50% for lower-tier SOC tasks, freeing human analysts to focus on more advanced investigations and response activities, Sordilla says.
Reducing alert fatigue
In alert triage, AI agents are reducing alert fatigue by clustering alert patterns and enabling risk-based prioritization, adds Dipto Chakravarty, chief product and technology officer at Black Duck.
For example, natural language processing agents can summarize threat alerts at scale and correlate them with threat intel feeds such as CVE.org and the CISA KEV Catalog, he says.
“The general incident response workflow is one of the beneficiaries of AI agents where we are seeing the value of automated playbooks for common incidents,” he notes.
AI agents are also playing a role in enriching threat intelligence at scale by ingesting and correlating threat intel from myriad sources and consequently enriching these alerts with value-added context such as CVE data.
“AI agents today can effectively accelerate derivation of insights from organized and normalized datasets,” by allowing analysts to ask questions in natural language, says Nicole Bucala, CEO at Databee. They eliminate the need for the specialized queries, analytical dashboards, or manual analysis typically required for the task.
Instead of flooding analysts with thousands of low-confidence warnings, AI-enabled detection platforms can score and correlate alerts, group related activity into higher-fidelity incidents, and filter out routine or benign behavior. The result, vendors and analysts say, is a reduction in alert fatigue and a shift in analyst workflows away from manual triage toward deeper investigation and response.
“AI is helping SOCs escape ‘activity theater’ by turning raw noise into faster, higher-confidence decisions backed by evidence,” says Craig Jones, chief security officer at Ontinue.
SOC burnout is a real concern, Jones notes. The biggest drivers of this in the industry are alert volume, fragmentation, and ambiguity, and those pressures exist for any team operating at scale. Analysts, he says, often end up spending too much of their day working high-alert loads that are low signal and then having to context-switch across multiple tools just to assemble the basics of an investigation.
Containing threats sooner
The real win with AI isn’t processing more alerts or closing more tickets; it’s about containing real threats sooner, with fewer mistakes, Jones says.
“When AI is used to correlate weak signals into coherent incidents, enrich investigations automatically, and recommend safe next actions inside clear guardrails, you stop measuring effort and start proving outcomes,” he explains.
Security experts expect AI to change the skills needed in security teams. Rather than eliminating jobs, it will help security teams automate routine tasks and shift roles toward engineering and system design, Accenture’s Sordilla says. The traditional SOC analyst role — focused heavily on manual log review — is likely to evolve into security engineering roles focused on building resilient systems, automation pipelines, and AI-assisted defenses.
Early data shows organizations that have deployed AI for detection engineering are seeing some measurable gains. In a Google study of 3,466 senior leaders, nearly seven in ten (67%) early adopters of agentic AI reported seeing it having a positive impact on their security posture. Of this group, 85% reported described AI as having improved their ability to identify threats. Early adopters of AI, Google noted, are seeing quantifiable benefits not just in terms of efficiency, but also in terms of efficacy.
At the same time, experts caution that AI-driven detection is not a silver bullet. Adversaries are increasingly experimenting with AI themselves — using it to generate more convincing phishing campaigns, automate reconnaissance, or modify malware to evade signature-based defenses. That dynamic is pushing defenders to treat AI not simply as another security tool, but as part of a broader evolution in security operations where human expertise, threat intelligence, and machine learning must work together.
“Cyberattacks have been industrialized at machine speed,” says Ram Varadarajan, CEO at Acalvio “We need to respond in kind.”
That means implementing defensive AI that can handle high-volume technical tasks such as triaging phishing emails, analyzing massive network logs for behavioral anomalies, deploying AI-aware cyber deception, and autonomously quarantining compromised endpoints to prevent lateral movement, he says.
“When it’s a machine-speed AI attacker, no human will ever be able to keep up, and these complex AI attacks are going to be launched at scale,” he notes.
Implementing AI correctly
The key to getting the most value out of AI in threat detection is to ensure humans are involved. Any threat finding or resulting remediation action based on those insights, especially those involving nontrivial consequence for business operations, should remain under human oversight, at minimum, says Databee’s Bucala. 
“Human in the loop is the mantra,” she says. “There’s a lot of business risk that can be incurred through full automation unless the margin of error in machine made decisions is close to zero.”
While AI shows promise in threat detection, it still needs refinement. The best practice for organizations is to establish a process that includes human validation, and humans who have the right attention to detail and context to spot check AI summary results and decisions, Bucala notes.
AI, adds Accenture’s Sordilla, is not a substitute for basic security hygiene. If an organization already has weak security practices, AI may simply accelerate existing problems. So, companies should first ensure they have strong governance, clear security standards, and mature processes — such as those outlined in frameworks from NIST and International Organization for Standardization — before layering AI into their security programs.
“AI is force multiplier,” Sordilla says. “If your company is heading in the wrong direction, you are going down the drain faster,” by deploying AI incorrectly, he cautions.
View the full article
AI is no longer a speculative topic for security leaders. It has moved from experimentation to implementation, and increasingly, to measurable production impact.
Over the past year, my conversations with CISOs have shifted. The question is no longer whether AI belongs in cybersecurity; it’s about deploying it responsibly, strategically and at scale.
For security leaders, this is not simply a technology decision. It is an operating model decision.
Organizations that treat AI as another layer added to existing workflows may see incremental efficiency gains. Those that treat it as an inflection point in how security operations conduct investigative work can fundamentally reshape their defensive posture.
From these CISO conversations, several realities are emerging that security leaders should confront directly.
The threat has accelerated beyond human scale
Recent threat intelligence underscores the urgency. CrowdStrike’s 2026 Global Threat Report found an 89% year-over-year increase in AI-enabled adversary activity. More concerning than volume is velocity. The average eCrime breakout time — the interval between initial compromise and lateral movement — dropped to 29 minutes, with the fastest observed breakout occurring in 27 seconds.
In one documented intrusion, an attacker gained access, moved laterally and began data exfiltration within four minutes.
These timelines compress the window for detection and response to a degree that challenges human-only workflows. Manual triage and sequential investigation processes struggle to keep pace with machine-speed attacks.
This is a material shift in tempo, and it requires a corresponding shift in defensive capability.
The questions have matured
The AI discussion in security has evolved in phases.
First came skepticism from security leaders, asking whether AI actually works in security operations. Given years of overpromised technology, the caution was warranted.
Experimentation followed, with questions centering on what types of work AI should handle and where it introduces risk.
Now, the dominant questions are more operational:
How do we deploy AI into production SOC workflows? How do we implement it quickly without disrupting already strained teams? What should our analysts focus on once AI absorbs repetitive tasks? These go beyond theoretical considerations and reflect a recognition that AI has crossed from possibility to implementation.
The cyber AI parity window
Historically, offensive cyber capabilities have benefited from asymmetry. Nation-state actors often developed and deployed advanced capabilities years before defenders became aware of them. By the time tools were exposed or leaked, adversaries had already compounded their advantage.
AI represents a break from that pattern.
The same foundational AI advancements powering offensive capabilities are also enabling a new generation of defensive tools. Unlike prior technological shifts, AI was not restricted to classified environments for years before becoming commercially available. It emerged publicly and broadly.
For the first time, defenders and adversaries gained access to a transformative technology at roughly the same moment.
This creates what I call the Cyber AI Parity Window — a limited period during which defenders are not structurally behind in technological capability.
Parity, however, is not the same as advantage. Advantage accrues to those who operationalize AI most effectively and most quickly.
This window will not remain open indefinitely.
Architecture determines scalability
Early enthusiasm around large language models led some to assume that a single, powerful AI system could manage security investigations end to end. Production deployments revealed the limits of that approach.
Security investigations are rarely linear. They involve contextual interpretation, cross-tool correlation, iterative reasoning and validation. Single-agent systems often struggle to sustain accuracy under these conditions.
More effective deployments rely on coordinated, multi-agent architectures. Specialized agents handle enrichment, reasoning, validation and response orchestration, dynamically adapting to alert type and environment.
While this architecture is more complex, it has proven more reliable at scale.
For CISOs, architectural transparency should be a priority. Understanding how systems reason, manage ambiguity and maintain accuracy under load is essential. In security operations, reliability is a requirement, not a feature.
Context is the control plane
Another consistent lesson from early deployments is that AI performance is inseparable from contextual depth.
Generic AI models cannot accurately investigate security events without understanding the environment they are protecting. Network architecture, identity models, detection logic, asset criticality and business workflows all shape investigative conclusions.
As organizations assign greater responsibility to autonomous systems, contextual misalignment can introduce risk rather than reduce it.
Successful implementations treat context as infrastructure. AI systems are deeply integrated with telemetry sources and workflows. Data pipelines are structured deliberately. Environmental fidelity is treated as foundational.
AI only amplifies the importance of understanding your environment.
From execution to management
Public discourse often frames AI in terms of job displacement. Within security organizations, the more relevant discussion is about redefining roles.
Security teams face persistent growth in alerts and talent shortages. Analysts spend significant time on repetitive investigations that require diligence but not necessarily strategic judgment.
AI creates an opportunity to shift human contribution from execution to management.
Rather than manually triaging alerts, analysts can define investigative logic. Instead of performing routine enrichment tasks, they can determine escalation thresholds. Instead of executing playbooks, they can design and refine them.
This mirrors transitions seen in other industries as automation matured: human value moves upstream, toward oversight, design and improvement.
In organizations that implement this shift thoughtfully, teams report not only reduced backlog but also improved engagement. Analysts work on more complex problems and develop more strategic capabilities.
The central question, therefore, is whether AI elevates the way expertise is applied, not whether it reduces headcount.
The window requires action
Defenders possess structural advantages that attackers do not. Large technology providers process trillions of security signals daily. Empirical research, including IBM’s Cost of a Data Breach Report, shows that organizations extensively using AI and automation experience lower breach costs and faster containment times.
But structural advantage compounds only with execution.
Every month that security operations remain dependent on manual triage is a month in which AI-enabled adversaries continue to optimize their workflows. The acceleration in breakout times does not pause for budget cycles or extended vendor evaluations.
The Cyber AI Parity Window represents a rare strategic opportunity. For once, defenders are not reacting to a capability that adversaries monopolized for years.
The question is whether organizations will capitalize on that parity before it narrows.
Production metrics over vision
Security leaders today evaluate AI platforms with appropriate rigor. Claims of transformative capability are insufficient.
Several standard operational metrics matter:
Investigations completed autonomously Average investigation time False positive and false negative rates Percentage of cases requiring human override Time to deployment and value realization AI must demonstrate measurable performance in production environments. Trust is built through documented outcomes, not conceptual promise.
Leadership in the AI production era
AI in cybersecurity represents a structural shift in how investigative work is conducted and how human expertise is applied.
CISOs now face a consequential choice: layer AI incrementally onto existing workflows or integrate it as a foundational component of security operations.
Organizations that succeed will demand measurable production outcomes, invest in contextual integration, evaluate architectural robustness, redesign workflows to elevate human expertise and act before the Cyber AI Parity Window closes.
The industry has moved beyond experimentation. AI is operating in production. Adversaries are leveraging it at machine speed.
The inflection point has arrived. What follows depends on execution.
This article is published as part of the Foundry Expert Contributor Network.
Want to join?
View the full article
Cybersecurity researchers have discovered a new campaign in which a cluster of 108 Google Chrome extensions has been found to communicate with the same command-and-control (C2) infrastructure with the goal of collecting user data and enabling browser-level abuse by injecting ads and arbitrary JavaScript code into every web page visited. According to Socket, the extensions are publishedView the full article
Mdisk – shutterstock.com
Hybride Attacken auf kritische Infrastruktur in Deutschland und Bundeswehr-Truppen im Ausland nehmen weiter zu. Spätestens seit 2022 sei ein spürbarer Zuwachs zu verzeichnen, sagte der Bundeswehr-Inspekteur Cyber- und Informationsraum, Vizeadmiral Thomas Daum, bei einem Pressetermin bei der Nato-Cyberabwehrübung «Locked Shields» im niederrheinischen Kalkar. 
Cyber-Angriffe gegen die Bundeswehr richteten sich gegen Rechenzentren in Deutschland oder gegen Truppen im Ausland – etwa in Litauen, wo Stationierte den Eindruck haben, dass sie am Telefon abgehört werden. In Litauen seien auch Desinformationskampagnen gestartet worden – zum Beispiel mit der falschen Behauptung, dass der Kontingentführer am Wochenende in Moskau gewesen wäre, um sich mit Freunden zu treffen. 
Größte Übung zur Cybersicherheit
Datenschützer registrieren seit Jahren Drohnen-Sichtungen und Ausspähversuche, Eindringversuche in Kasernen, die Beschädigung von Datenkabeln und Versorgungsleitungen in der Ostsee sowie Störungen des GPS-Systems. Die mutmaßlichen Hauptverursacher säßen in Russland, China, Iran und Nordkorea, sagte ein Bundeswehrsprecher in Kalkar.
Die Cybertruppe der Bundeswehr beteiligt sich mit rund 40 anderen Nationen an «Locked Shields». Die Unternehmung wird aktuell als weltweit größte und komplexeste multinationale Übung der Nato zur Cybersicherheit bezeichnet. Es werden Echtzeitangriffe eines «Teams Rot» simuliert, die von einem «Team Blau» abgewehrt werden müssen. 
Aus Deutschland sind auch Polizei- und Sicherheitsbehörden und zivile IT-Fachleute etwa von der Telekom beteiligt.
Energieversorger, Banken – und das Einwohnermeldeamt?
Zur kritischen Infrastruktur zählen in Deutschland große Energieversorger, Banken und IT-Dienstleister – nach Schätzungen mehr als 29.000 Unternehmen. In den Fokus könnten aber auch Behörden geraten, sagte ein Bundeswehrsprecher. Russland habe beispielsweise kurz vor dem Angriff auf die Ukraine die dortigen Einwohnermeldeämter attackiert, um deren Daten zu löschen und so eine Mobilisierung der Bevölkerung zu erschweren. Die Ukraine habe die Daten aber rechtzeitig gesichert. (dpa/ad)
View the full article
A critical security vulnerability impacting ShowDoc, a document management and collaboration service popular in China, has come under active exploitation in the wild. The vulnerability in question is CVE-2025-0520 (aka CNVD-2020-26585), which carries a CVSS score of 9.4 out of 10.0. It relates to a case of unrestricted file upload that stems from improper validation ofView the full article
The U.S. Cybersecurity and Infrastructure Security Agency (CISA) on Monday added half a dozen security flaws to its Known Exploited Vulnerabilities (KEV) catalog, citing evidence of active exploitation. The list of vulnerabilities is as follows - CVE-2026-21643 (CVSS score: 9.1) -  An SQL injection vulnerability in  Fortinet FortiClient EMS that could allow an unauthenticated attacker toView the full article
Over the past week, reaction to Anthropic’s Glasswing disclosure has split along familiar lines. At one end: alarm over an AI system capable of autonomously identifying and exploiting vulnerabilities. At the other: dismissive hot takes, arguing there is nothing new here.
A more grounded view comes from a new briefing by the Cloud Security Alliance (CSA), led by Gadi Evron, CEO of Knostic and CISO-in-Residence for AI at the alliance; Rob T. Lee, chief AI officer and chief of research at SANS Institute; and Rich Mogull, chief analyst at CSA.
The paper draws on a deep bench of contributors, including former CISA Director Jen Easterly, Bruce Schneier, former National Cyber Director Chris Inglis, and former Google CISO Phil Venables, along with dozens of CISOs and CEOs.
Evron told CSO that assembling that level of input among so many leaders so quickly reflects the nature of cybersecurity itself: “The cybersecurity industry is also a community, and knowing each other, all folks need to have is a good cause, and dispelling noise and spreading good information matters to us.”
The group’s conclusion is direct: Glasswing is not an outlier. It is an early example of a capability that will scale, and CISOs should start getting ready for this era.
“In the near term, security organizations will likely be overwhelmed by the need to apply patches and respond to AI-discovered vulnerabilities, exploits, and autonomous attacks,” the paper states. “The storm of vulnerability disclosures from Project Glasswing is the first of many large waves.”
The shift is speed
AI-driven vulnerability discovery is not new. What has changed is speed. Tasks that once took weeks or months — finding a flaw, building an exploit, chaining it into an attack — can now happen in hours.
According to the paper, “Anthropic’s Claude Mythos (Preview) represents a step change in that trajectory, autonomously finding thousands of critical vulnerabilities across every major operating system and browser, generating working exploits without human guidance, and empowering autonomous attack orchestration, all at a speed and scale that outpaces any prior capability.”
This acceleration deepens a familiar asymmetry: Defenders must be right consistently, whereas attackers only need to succeed once.
Moreover, “The window between discovery and weaponization has collapsed to hours. Attackers gain disproportionate benefit, and current patch cycles, response processes, and risk metrics were not built for this environment,” the paper states.
“Building a ‘Mythos-ready’ security program is not about reacting to one model or announcement. It is about permanently closing the gap between how fast vulnerabilities are found and how fast your organization can respond.”
Claude Mythos Preview is a step up
A separate analysis from the UK’s AI Security Institute (AISI) evaluated Mythos Preview itself.
The evaluations involved both capture-the-flag (CTF) challenges and more complex ranges designed to simulate multi-step attack scenarios, where the model outperformed other AI systems.
Mythos Preview came out on top in a 32-step corporate network attack simulation spanning initial reconnaissance through to full network takeover, which the Institute estimates requires humans 20 hours to complete.
AISI’s tests also showed that Mythos Preview is capable of autonomously attacking small, weakly defended enterprise systems once access is obtained. “Our testing shows that Mythos Preview can exploit systems with weak security posture, and more models with these capabilities will likely be developed,” AISI concluded.
What CISOs should do now
AISI’s recommendation to organizations is that they should strengthen fundamentals, including regular application of security updates, robust access controls, security configuration, and comprehensive logging.
It advises, “Future frontier models will be more capable still, so investment now in cyber defence is vital. AI cyber capabilities are dual use; while they pose security challenges, they can also help deliver game-changing improvements in defence.”
The CSA paper highlights three predictions for CISOs.
Operationally: Expect a surge of patches from the approximate 40 vendors in the early access program, potentially mirroring recent periods where multiple supply chain incidents required response within a two-week window.
Risk management: Business risk is shifting, requiring close engagement with stakeholders on risk planning and tolerance. The CISO’s ability to manage risk is becoming more constrained, with potential downstream effects on reporting and projections.
Strategically: Conduct longer-term gap analysis and selectively overhaul key functions, including governance processes that enable faster technology onboarding and the deployment of AI-driven security controls.
The report also elevates Mythos to a board-level issue, allowing CISOs to frame current capabilities and make the case for further investment.
The bottom line, as the CSA paper concludes, is that “AI-based attacks represent a structural shift in how offense and defense work, and it will not change. The cost and capability floor to exploit discovery is dropping, the time between disclosure and weaponization is compressing toward zero, and capabilities that previously required nation-state resources are now becoming broadly accessible.”
See also: “Cybersecurity in the age of instant software”
View the full article
Production on the iPhone Fold is behind schedule, but the device is still slated to launch in fall 2026, reports DigiTimes. The site says that production has been pushed back by "roughly one to two months," but Apple has not communicated any launch delays to suppliers. Apple is still planning for a 2026 launch, which suggests a tighter production schedule.


Apple planned to begin ‌iPhone Fold‌ mass production in June 2026, but mass production has now slipped to early August. Some rumors have suggested that Apple is having more manufacturing problems than expected with the Engineering Validation Test phase that the ‌iPhone Fold‌ is in right now, but other rumors suggest that won't lead to delays.

Last week, Japanese site Nikkei said that the engineering delays could cause Apple to delay the launch of the ‌iPhone Fold‌ until 2027, but Bloomberg's Mark Gurman said the ‌iPhone Fold‌ remains on track to launch "around the same time" or "soon after" the iPhone 18 Pro models in September 2026.

DigiTimes' report reconciles the two separate launch timing rumors, indicating that there is indeed a delay in the test phase, but Apple may be able to make up ground during the mass production phase. Prior to mass production, the ‌iPhone Fold‌ will still need to go through Design Validation Testing and Production Validation Testing.

A delay in the mass production timeline could cause serious supply chain shortages, and ‌iPhone Fold‌ availability is likely to be constrained. When we hear rumors about launch issues this early in an iPhone cycle, it typically leads to severe launch shortages and a device that sells out in minutes during pre-orders.

The ‌iPhone Fold‌ is expected to cost between $2,000 and $2,500. It will be Apple's first foldable device, with a 5.5-inch display when closed and a 7.8-inch display when open. More about the device can be found in our iPhone Fold roundup.Related Roundup: iPhone FoldTags: DigiTimes, Foldable iPhone
This article, "iPhone Fold Production Pushed Back, But Fall 2026 Launch Still on Track" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
American Airlines is now supporting the enhanced boarding pass feature that Apple added in iOS 26. The app's latest update says boarding passes in the Wallet app feature an updated look with Apple Maps integration, destination guides, and luggage tracking capabilities.


Apple revamped boarding passes in ‌iOS 26‌ to make them more useful to passengers who are using their iPhones for flights in lieu of a paper boarding pass. Boarding passes have maps for navigating through airports, an option to tap into Find My to track items equipped with AirTags, and tools for reporting missing bags to airlines. From American Airlines:

Airlines are need to opt in to the new boarding passes and update their apps with support. United Airlines, Air Canada, and Southwest already have the enhanced boarding passes. Delta briefly added support, but removed it shortly after.

Apple also implemented Live Activities support for airlines so users can get real-time updates on a flight in ‌iOS 26‌, but it is not clear if American Airlines is supporting that feature based on the app's release notes.

Back in June 2025, Apple said American Airlines was one of the airlines that would add support for the refreshed boarding passes. Other airlines listed included Air Canada, Delta Air Lines, JetBlue, Jetstar, Lufthansa Group, Qantas, Southwest Airlines, United Airlines, and Virgin Australia.

American Airlines has also supported the ‌Find My‌ lost luggage tracking feature since February 2025. In iOS 18, Apple worked with airlines to make it easier for airline employees to locate lost baggage using AirTags.Related Roundups: iOS 26, iPadOS 26Tag: American AirlinesRelated Forum: iOS 26
This article, "American Airlines Now Supports iOS 26's Revamped Wallet Boarding Passes" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
With the second beta of iOS 26.5, Apple is continuing to prepare for ads in the Apple Maps app. There's a new splash screen in the app that says it will display ads based on approximate location, current search terms, or a view of the map when searching. Ads will also be shown in the "Suggested Places" section that was added in the first beta.


Apple says that advertising information is not linked to an Apple Account to preserve user privacy, so ads viewed and interacted with will not be associated with an Apple Account. Data is not collected or stored by Apple and not shared with third parties, according to Apple.

Signs of ads in the Maps app were spotted in the first beta, but iOS 26.5 testers could start seeing ads with the second beta. It's not yet clear if Apple has just implemented the splash screen, or if ads are going to begin showing up in the near future.


Apple announced plans to introduce ads for the Maps app back in March, with ads set to roll out to the public "this summer." Apple plans to include ads in the Maps app on iPhone, iPad, and Mac in the United States and Canada.

Ads in ‌Apple Maps‌ will have an "Ad" label, much like ads in the App Store.Related Roundups: iOS 26, iPadOS 26Tag: Apple MapsRelated Forum: iOS 26
This article, "Apple Maps Ads Move Closer to Launch With iOS 26.5 Beta 2" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Amazon this week has all-time low prices on the Apple Watch Series 11, with $100 discounts across numerous models of the smartwatch. This sale includes nearly every aluminum model of the Series 11 on sale at a record low price.

Note: MacRumors is an affiliate partner with Amazon. When you click a link and make a purchase, we may receive a small payment, which helps us keep the site running.

You can get the 42mm GPS Apple Watch Series 11 for $299.00, down from $399.00, and the 46mm GPS model for $329.00, down from $429.00. On Amazon, you'll find four of both the 42mm and 46mm GPS models on sale at these all-time low prices.

$100 OFFApple Watch Series 11 (42mm GPS) for $299.00
$100 OFFApple Watch Series 11 (46mm GPS) for $329.00

If you're shopping for cellular models, you can find record low prices on multiple models this week on Amazon. The 42mm cellular Apple Watch Series 11 has hit $399.00, down from $499.00, and the 46mm cellular model has hit $429.00, down from $529.00.

$100 OFFApple Watch Series 11 (42mm Cell) for $399.00
$100 OFFApple Watch Series 11 (46mm Cell) for $429.00

Head to our full Deals Roundup to get caught up with all of the latest deals and discounts that we've been tracking over the past week.



Deals Newsletter

Interested in hearing more about the best deals you can find in 2026? Sign up for our Deals Newsletter and we'll keep you updated so you don't miss the biggest deals of the season!




Related Roundup: Apple Deals
This article, "Apple Watch Series 11 Hits $100 Off Nearly Every Aluminum Model on Amazon" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Banks and financial institutions in Latin American countries like Brazil and Mexico have continued to be the target of a malware family called JanelaRAT. A modified version of BX RAT, JanelaRAT is known to steal financial and cryptocurrency data associated with specific financial entities, as well as track mouse inputs, log keystrokes, take screenshots, and collect system metadata. "One of theView the full article
Apple today seeded the second betas of upcoming iOS 26.5 and iPadOS 26.5 updates to developers for testing purposes, with the software coming two weeks after Apple released updated first betas.


Registered developers can download the betas from the Settings app on the iPhone or iPad by going to the General section and selecting Software Update.

iOS 26.5 and iPadOS 26.5 do not include new Siri capabilities, suggesting any ‌Siri‌ updates are being held until iOS 27. The Maps app has a Suggested Places feature for recommending locations to visit nearby based on trends and recent searches, plus Apple is laying the groundwork for ads in the Apple Maps app.

Apple is continuing to test end-to-end encryption (E2EE) for RCS messages between iPhone and Android users. Apple included the feature in the iOS 26.4 beta, but removed it before the update launched to the public.

In the European Union, Apple is testing proximity pairing, notification forwarding, and Live Activities for third-party wearables like earbuds and smartwatches. The functionality will allow third-party wearables to have many of the same features as the Apple Watch and AirPods.

More detail on what's new in iOS 26.5 can be found in our iOS 26.5 beta features guide.Related Roundups: iOS 26, iPadOS 26Related Forum: iOS 26
This article, "Apple Seeds Second iOS 26.5 and iPadOS 26.5 Betas to Developers" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
Apple today provided the second beta of an upcoming macOS Tahoe 26.5 update to developers for testing purposes, with the update coming two weeks after the first beta.


Developers can download the ‌macOS Tahoe‌ 26.5 update by opening up the System Settings app, selecting the General category, and then choosing Software Update. Beta Updates will need to be enabled, and a free developer account is required.

No new features were found in the first ‌macOS Tahoe‌ 26.5 beta, and it's likely the update primarily focuses on bug fixes and performance improvements.Related Roundup: macOS TahoeRelated Forum: macOS Tahoe
This article, "Second macOS Tahoe 26.5 Beta Now Available for Developers" first appeared on MacRumors.com

Discuss this article in our forums

View the full article
OHC_logo_transparent_01.jpeg flags-medium.png OHC_logo_blue_square_small.jpeg

 

Account

Navigation

Search

Search

Configure browser push notifications

Chrome (Android)
  1. Tap the lock icon next to the address bar.
  2. Tap Permissions → Notifications.
  3. Adjust your preference.
Chrome (Desktop)
  1. Click the padlock icon in the address bar.
  2. Select Site settings.
  3. Find Notifications and adjust your preference.