California Governor Gavin Newsom unveiled a proposal on November 18, 2026, to strengthen state oversight of the artificial intelligence industry. It would authorize the state to order emergency shutdowns of dangerous AI systems.
The framework also calls for mandatory, ongoing safety testing by independent third-party auditors at every stage of development. It would give the state new powers to cap an AI system’s ability to create more powerful successor models.
The proposals build on California’s existing AI safety laws and are positioned as a national model in the absence of federal regulation. Developers would have to create verified protocols for shutting down their models; the proposed state authority to compel a shutdown is a separate power. Newsom urged the federal government to adopt the framework as national rules.