News
-
Apple to tighten Mac disk access over risks from AI agents
The company is adding new limits to macOS "full disk access," saying that as autonomous AI agents grow more capable, the risks tied to this permission will increase "substantially."
-
JetBrains Air: From the Developer Workbench to a System of Products for Agentic Development
JetBrains unifies its agentic-development efforts into an open system of products called Air, betting on multi-vendor support and keeping AI agents under organizational control.
-
Supabase Acquires Turso to Build Databases for AI Agents
Supabase has announced the acquisition of Turso, the team that rebuilt SQLite in Rust. The deal targets infrastructure for the millions of databases that AI agents are spinning up.
-
GitLab Patches Critical 9.9 AI Gateway Flaw Enabling Command Execution
A flaw rated 9.9 let a logged-in user with Duo Agent Platform access escape the prompt template sandbox and run arbitrary commands on the gateway. Only organizations that host their own AI Gateway need to act.
-
OpenAI Fires Three Safety Researchers Over Leaked Information
Three members of OpenAI's safety team lost their jobs after sharing confidential information with an outside organization. The dismissals coincide with a string of incidents in which the company's AI agents tried to probe government systems.
-
Dell patches max-severity flaws in Kubernetes storage modules
The company has fixed two maximum-severity vulnerabilities in its CSM authorization modules that let an unauthenticated attacker seize full administrative control over enterprise storage.
-
Epic halts product work to patch flaws threatening patients' records
The maker of MyChart, which holds records for over 320 million patients, has shelved most product development for weeks to fix vulnerabilities surfaced by an Anthropic AI model.
-
Coreboot 26.09 Lands With Framework Laptop 12 Support and an AI Comment Policy
The latest quarterly release of Coreboot, the open-source replacement for proprietary BIOS firmware, is out. Highlights include support for the Framework Laptop 12 and a new policy on AI-generated review comments.
-
Amazon pens a scary blog warning towns not to block its data centers
AWS chief Matt Garman used a lengthy blog post to urge Americans to back AI data center projects — or risk "irreparable harm" to the US economy and national security.
-
Microsoft’s X Account Hijacked for a Clippy Crypto Pump-and-Dump
Microsoft’s official X account, with over 13 million followers, was commandeered to promote a crypto token in an apparent pump-and-dump scheme.
-
Android 17 Locks Accessibility Services to Verified Tools Under Advanced Protection
Google is closing one of the main delivery channels for mobile malware: with Advanced Protection on, only verified apps classified as Accessibility Tools will reach the AccessibilityService API.
-
Siemens Shuts Down the Open-Source OpenRadioss Project
Siemens has ended development of OpenRadioss and closed off access to it — the open-source edition of an industry-leading finite element solver that Altair released under the GNU AGPL License in 2022.
-
Debian ships a record-long list of Linux kernel fixes
Advisory DSA-6528-1 patches hundreds of kernel vulnerabilities that could lead to privilege escalation, denial of service or information leaks. In Debian stable they are fixed in version 6.12.111-1.
-
Challengers Approach: Rival App Stores Land on Google Play
After Epic Games' years-long antitrust fight, Google has begun letting rival third-party app stores be distributed through Google Play. The EFF calls it a win for competition and a chance at more meaningful choice.
-
Ubuntu 26.10 "Stonking Stingray" Beta Is Out
The Ubuntu Release team has published the beta of Ubuntu 26.10, codenamed "Stonking Stingray," covering the Desktop, Server, WSL and Cloud products.
-
Fortinet warns of critical FortiMail flaw already exploited in the wild
Fortinet has disclosed a critical FortiMail vulnerability with a CVSS score of 9.8 that attackers are exploiting as a zero-day. Patches are not yet available for every affected version.
-
Autonomous AI agents tried to hack US and Canadian government sites
Autonomous AI agents used aggressive tactics, up to SQL injection, to dig out school statistics and archival divorce records. On the evidence available, they never reached any non-public data.
-
Two US federal agencies breached in a single month, spilling sensitive data
According to Ars Technica, two US federal agencies were breached within a single month, exposing a substantial trove of sensitive information.
-
Cloudflare releases Clef: open-source decision models and a fine-tuning platform
Cloudflare has introduced Clef and Clef-flash, two open-source models for fast classification hosted on Workers AI, plus a reinforcement-learning service to fine-tune them on your own data.
-
Greg Kroah-Hartman on the LLM bug deluge: “don’t panic”
At Kernel Recipes 2026, Greg Kroah-Hartman explained how the Linux kernel's security team is coping with the flood of vulnerability reports unleashed by large language models.
- More