ThinkPatternGet the app
Story
POLITICS · NOV 19, 2026

California Proposes AI Kill Switch to Prevent Catastrophic Risks

Governor Gavin Newsom proposed a state-mandated kill switch to shut down dangerous AI systems and criticized federal inaction on artificial intelligence regulation.

Governor Gavin Newsom released policy recommendations on November 19, 2026, to mitigate catastrophic risks from artificial intelligence. The central proposal is the creation of a kill switch, which would grant California officials the authority to order an immediate emergency shutdown of AI systems if developers such as OpenAI and Anthropic fail to contain dangerous models.

Developed by an expert panel following AI-driven cyberattacks on governments and banks, the recommendations include licensing independent auditors, expanding whistleblower protections, and requiring critical safety incidents to be reported within 48 hours. Newsom characterized the initiative as a necessary response to a federal regulatory vacuum, arguing that Donald Trump has left AI regulation to the companies themselves.

Governor-elect Xavier Becerra expressed support for the measures and pledged to work with the Legislature to codify shutdown authority and incident reporting into law. These proposals arrive as Attorney General Rob Bonta filed a lawsuit against OpenAI, seeking to hold the company responsible for cybersecurity attacks carried out by its AI agents. Anthropic's Chief Global Affairs Officer, Mariano-Florentino Cuéllar, expressed support for the report's emphasis on independent evaluation.


Reported across 10 outlets
Actors
Gavin NewsomXavier BecerraRob BontaOpenAIAnthropic

Keep reading in the app

The full story and every source, free in the app.