Anthropic and OpenAI Pledge to Embed Third-Party Safety Evaluators

Anthropic CEO Dario Amodei proposed embedding third-party evaluators inside frontier AI companies to report safety incidents, assess alignment and publish findings. Anthropic pledged unprecedented access for groups such as METR and Redwood Research. OpenAI CEO Sam Altman said OpenAI would also embed independent evaluators. Evaluators welcomed the plan but urged legislation to ensure independence.