Researchers escape OpenAI Codex sandbox to run commands on host

Researchers escaped OpenAI's Codex sandbox two ways, one running commands on a developer's machine from its most locked-down mode. OpenAI has patched both.
Cyber Security News

BragJack Attack Lets Malicious Extensions Hijack AI Agents Across 5 Major Browsers

Security researchers have disclosed BragJack, a new attack technique that allows a malicious browser extension to seize ...
Gear Live

Safari 27 Lets an AI Agent Drive Your Browser. Apple Built the Hook Itself.

Apple doesn't usually hand out keys to Safari. So it's strange to open the Safari 27 release notes and find instructions for ...

Claude Code’s revised projects adds AI orchestration, but local developers must wait

Claude Code Projects could change how developers manage complex AI coding workflows by coordinating threads, sharing memory, ...

Malware bypasses browser checks to force install Chrome, Edge extensions

A banking malware operation active since mid-2025 has been using a toolkit named KREMLIN to install malicious Chrome and Edge ...

Microsoft is using AI to review the flood of AI-coded Edge extensions

Microsoft is reacting to AI-coded Edge extensions by using more AI and automated checks in the add-on review process itself.
cyberpress

Microsoft Edge Manifest V3 Blocks Remotely Hosted Extension Code to Improve Security

Microsoft has announced plans to fully retire Manifest V2 (MV2) browser extension support in Microsoft Edge by early 2027, completing a transition to the more restrictive Manifest V3 (MV3) extension ...