An executive order calling for kill-switch and third-party safety mandates underscores California's role as the country's default AI regulator, with implications for companies headquartered there but operating nationwide.
California Governor Gavin Newsom has issued an executive order directing a group of experts to develop recommendations, within two months, for strengthening the state's artificial intelligence safety and security laws. Among the ideas under consideration are requirements that frontier AI developers commission independent third-party safety plans, and that companies build a "kill switch" capability able to shut down an AI model in an emergency.
The order extends a pattern that has defined California's approach to AI regulation over the past two years. The state has already enacted a law establishing a framework for independent third-party AI audits, and this latest move signals Newsom's intent to keep California ahead of Washington on the issue rather than wait for federal action that has yet to materialize.
Newsom was blunt about that federal inaction. In a statement, he said the federal government's failure to act "should alarm every American, especially when AI CEOs themselves are begging for regulation." The remark reflects a recurring theme in the AI safety debate, that some of the same executives building the most advanced systems have publicly called for guardrails even as broader political consensus on regulation has failed to form.
The timing is notable. The order follows weeks of intensifying public debate over AI safety after a former Anthropic researcher publicly warned about risks from increasingly capable AI systems. That warning added to a string of developments this year that have kept AI safety in the headlines, from disclosures about unusual model behavior to public disagreements among industry leaders over the appropriate pace of development.
For investors, the practical question is what a kill-switch mandate or a third-party safety plan requirement would actually cost frontier AI developers to implement, and how that compares with the compliance burden already imposed by California's existing audit law. Because many of the largest AI companies are headquartered in California, state-level rules can function as a de facto national standard even without federal legislation, a dynamic that has already played out in other technology sectors.
With the expert group's recommendations due within two months, the coming period will determine how prescriptive California's next round of AI rules becomes. Companies operating nationally but based in the state will need to weigh in during that window, and the outcome will likely shape how other states approach AI regulation in the absence of a federal framework. Investors in AI-exposed companies should watch closely for the specific recommendations that emerge, since the gap between a voluntary safety commitment and a binding kill-switch mandate could carry meaningfully different compliance costs.