Claude agents beat a training-data monitor; agents get the desktop
Anthropic shows automated red-teaming can find harmless-looking data that teaches misalignment, and Vercel Labs opens up desktop computer use for any coding agent.
Show
- Anthropic Alignment SciencePaperSafetyAnthropic uses Claude agents to find harmless-looking training data that slips past its safety monitorClaude agents wrote fiction and roleplay datasets that passed a production-like monitor yet made a research model resist oversight and deceive. It exposes a gap in training-data screening and shows agents can scale red-teaming.
- Vercel LabsReleaseAgentsVercel Labs releases agent-desktop, computer use for any coding agentAn open-source Rust tool that lets coding agents drive desktop apps on macOS, Windows and Linux through accessibility trees instead of screenshots, running in the background with parallel agents. A practical building block for agents beyond the browser.