Anthropic and OpenAI need truly independent safety evaluators, experts say in public letter

Over 100 AI experts, including Geoffrey Hinton and members from Johns Hopkins and Stanford, urge AI companies like Anthropic and OpenAI to grant independent evaluators full access and protections to rigorously test AI safety. This coalition seeks standardized principles to ensure transparency, objectivity, and accountability in AI safety assessments amid growing concerns about frontier models. Anthropic CEO Dario Amodei has proposed giving evaluators enhanced access to inspect AI models, aiming,