Ex-OpenAI safety lead challenges Anthropic's claim of understanding model risks

Former OpenAI safety researcher Miles Brundage criticized Anthropic's claim that it largely understands and can manage current model risks as clearly inaccurate, adding that one can warn things could get worse without overstating control over risks.

2026-09-25 ~ 2026-09-25 · 2 related posts

• Ex-OpenAI safety lead Miles Brundage calls Anthropic's 'we largely understand model risks' claim obviously false — Miles_Brundage · 2026-09-25
• Brundage follows up: Anthropic can contrast present vs future risks without claiming we're on top of them — Miles_Brundage · 2026-09-25