Anthropic and OpenAI diverge on who should pull the plug on risky AI

Anthropic and OpenAI agree on the need for stronger oversight of frontier AI systems but differ on risk mitigation. Anthropic proposes empowering courts to suspend dangerous systems, whereas OpenAI suggests that developers must maintain effective shutdown controls. These proposals follow Governor Gavin Newsom's September 18 executive order regarding AI safeguards. Both companies support expanded incident reporting and independent safety assessments for AI systems that operate beyond their intended permissions to ensure overall safety and accountability.