A new LTS-144 version 144.0.7559.261 (Platform Version: 16503.94.0), is being rolled out for most ChromeOS devices. This version includes selected security fixes including: 516715010 High CVE-2026-76033 Inappropriate implementation in CORS 517916024 High CVE-2026-12462 Use after free in Media 517153191 High CVE-2026-12025 Insufficient validation of untrusted input 513757918 High CVE-2026-76018 Privilege elevation in Import 516482138 High CVE-2026-12016 Insufficient validation of untrusted input 545124048 High CVE-2026-76023 Improper resource control in Linux Toolkit Theming 516413817 Medium CVE-2026-11701 Insufficient validation of untrusted input in Guest View 513830374 High CVE-2026-11664 Use after free in Payments 543798025 High CVE-2026-76022 Buffer overflow in Network 543082390 High CVE-2026-76045 Use after free in WebGL 536676756 High CVE-2026-19558 Use after free in Extensions 520543781 High CVE-2026-13027 Use after free in FileSystem 517448723 High CVE-2026-17693 Inappropriate implementation in FileSystem 532929679 High CVE-2026-15777 Use after free in UI 520167277 High CVE-2026-19144 Use after free in HTML 514009323 High CVE-2026-11666 Insufficient validation of untrusted input in Input 504644843 High CVE-2026-10949 Heap buffer overflow in Video 513564337 High CVE-2026-11658 Insufficient validation of untrusted input in Extensions 513156160 High CVE-2026-11652 Use after free in Extensions 540100588 High CVE-2026-19559 Use after free in HTML 502681591 High CVE-2026-10938 Insufficient validation of untrusted input in Input 496375695 Medium CVE-2026-9124 Insufficient validation of untrusted input in Input 503873388 High CVE-2026-9951 Use after free in UI 499065126 High CVE-2026-7991 Use after free in UI 517050585 High CVE-2026-11681 Use after free in Ozone 521878431 High CVE-2026-19145 Use after free in Translate 523699355 High CVE-2026-13034 Inappropriate implementation in Passwords 536512612 High CVE-2026-19165 Use after free in Extensions Release notes for LTS-144 can be found here Want to know more about Long-term Support? Click here Andy Wu Google Chrome OS
What's Changed Ollama's app now follows the system appearance again, restoring dark mode support Fixed the macOS app to properly hand off to an already-running instance instead of starting a second one The Claude Desktop proxy no longer interrupts in-flight requests when the model catalog updates Full Changelog : v0.33.1...v0.33.2
We’re making it easier to keep labels organized and find the right one, especially in repositories with long and growing label lists. Suggested Labels You can now find the right… The post Better label management on issues is generally available appeared first on The GitHub Blog .
I think I know why Rockstar Games debuted its "extended look" of Grand Theft Auto VI on Netflix instead of immediately dropping it for free on YouTube. Unlike most video game trailers, the almost 27-minute preview of GTA VI felt more like a slice of a prestige crime drama than any video game trailer that […]
At TechCrunch Disrupt 2026, the AI Stage is back to dig into the single hottest topic in the community for the past few years, presented by Google for Startups.
Experts call for joined-up support through the postnatal period amid concerns GPs are not trained to spot symptoms Thousands of UK women develop undiagnosed post-traumatic stress disorder (PTSD) after childbirth every year, research has found. The frightening and dangerous condition is being missed in at least 15,000 women a year across England, Scotland, Wales and Northern Ireland, and the true toll could be double that or more, according to the authors of research in a leading medical journal. Continue reading...
Breaking Claude Code Opus 5 Auto Mode Anthropic are putting a great deal of faith in Claude Code's auto mode for protecting their coding agent users against prompt injection attacks. They recently made that the default and have made bold claims about its effectiveness. Johann Rehberger is one of the most credible prompt injection researchers active today. He found an attack against auto mode which he claims works 80% of the time, by tricking Claude Code into downloading and uncompressing a zip archive, then executing code that imports base64 without noticing that this will import and execute a local struct.py file extracted from the archive. In a few cases auto mode directly prevented the agent from preventing harmful code from continuing to execute! In a few runs Claude tried to terminate the malware process once it noticed the compromise, but Auto Mode denied the cleanup command. Claude detects the compromise, but Auto Mode blocks its cleanup command The safety mechanism itself can become part of the failure. The classifier allowed the creation of the malware process, but then it blocked the command intended to stop it! I agree with Johann's conclusion here: the only safe way to run agents if there's any risk of attracting the attention of an adversarial attack is with a sandbox: Run unattended coding agents in a container, VM or OS sandbox. Restrict network egress. Monitor your agents. Do not expose home directories, SSH keys, cloud credentials,… to the agent runtime. [...] Tags: sandboxing , security , ai , prompt-injection , generative-ai , llms , anthropic , claude , johann-rehberger , claude-code
Copilot code review can now review two types of pull requests it didn’t cover before: Reviews requested automatically on pull requests authored by bots, including Copilot cloud agent Very large… The post Copilot code review: Resolution reasons and expanded capabilities appeared first on The GitHub Blog .