Anthropic just opened its doors to outside safety watchdogs. Here's how it will work.

What happened: Anthropic is giving six independent safety evaluators from METR employee-like access to its hardware, office space, and internal tools to vet unreleased models and agents. It is the first time a frontier AI company has done so. Why it matters: The move fulfills a pledge CEO Dario Amodei made last month. OpenAI CEO Sam Altman said his company would also embed evaluators, but that has not happened yet. Key numbers: Six evaluators from METR.

Anthropic opens safety oversight to outside watchdogs, signaling AI governance catalyst.

Anthropic opens safety oversight to outside watchdogs, signaling AI governance catalyst. — impact context not separately stored; see What changed.