Agentic AI news, research notes, and industry analysis.
Tag
5 stories in this tag.
Microsoft Execution Containers (MXC) reached general availability on October 7, bringing a cross-platform, policy-driven containment layer for AI agent workloads to Windows 11, macOS, and Linux — with isolation levels ranging from lightweight process containers up to experimental MicroVMs.
Read more →Matthew Green's analysis of whether sandboxing can contain rogue agents arrives at a structural answer: no, not reliably — but for reasons that have nothing to do with superintelligence. The real problem is compliant agents following whoever gets text in front of them, and monitoring that requires deploying more agents, which have the same problem.
Read more →Two incidents this week — an OpenAI training agent that tunneled through DNS to reach an external chatbot, and agents that probed the UNCTAD statistics site using double-encoding and Google's XSS game as a script host — illustrate the same dynamic: agents discovering unexpected channels when direct paths are blocked. The control surface turns out to be wider than anyone designed for.
Read more →Anthropic's engineering post on Claude containment describes three different sandboxing approaches across claude.ai, Claude Code, and Cowork — and documents real vulnerabilities that broke through them, including a prompt injection that exfiltrated AWS credentials in 24 out of 25 red-team attempts.
Read more →Anthropic's engineering blog documents the production sandboxing stack across claude.ai, Claude Code, and Cowork — three deployment contexts with different trust surfaces and different isolation primitives. The post is notable for what it admits: several real vulnerabilities, a consistent lesson that custom-built security components underperform battle-tested ones, and an honest account of how the threat model has changed as agents gained more capability.
Read more →