California Moves to Lead AI Oversight with 'Kill Switch' Proposal
California Governor Gavin Newsom is positioning the state to take a leading role in AI oversight, issuing an executive order on Friday that directs state agencies to examine stronger safety requirements for advanced AI systems.
The order establishes a timeline for action: a group of experts will deliver recommendations within two months on how California could strengthen AI safety measures in state law. The group is being asked to consider several specific policy options.
Key areas under consideration include requiring AI companies to embed independent verification groups onsite for regular audits of their systems. The expert panel will also examine whether transparency reports and risk assessments from AI companies should be subject to standards set by independent auditors.
Perhaps most notably, the order directs the group to evaluate the feasibility of mandating a "kill switch" for frontier AI models—a mechanism that could shut down or disable an AI system under certain conditions.
The initiative reflects California's interest in establishing itself as a key regulatory voice in the rapidly evolving AI landscape, while seeking input from technical and policy experts on practical implementation approaches.