Newsom Signs AI Kill-Switch Order Before Term Ends

California Governor Gavin Newsom, whose final term ends in January, has signed executive orders in his last weeks as governor calling for independent verification of artificial intelligence models and setting up an expert committee to design kill switches for AI agents that could go rogue and cause harm, according to Al Jazeera. The move is the latest chapter in an aggressive state-level push for Newsom AI regulation as the governor seeks to shape how frontier AI systems are overseen before he leaves office.

The orders also expand the definition of AI safety incidents that developers must report, Al Jazeera reported on October 17. The directives revive provisions from a 2024 state assembly bill that Newsom vetoed at the time, saying the measures would amount to “thwarting the promise of this technology.”

Newsom AI regulation: What the Executive Orders Cover

The new orders contain three main elements:

• Independent verification of AI models: The state will seek outside review of the safety practices of frontier AI systems rather than relying solely on developers’ own assessments.
• A kill-switch committee: An expert committee will be tasked with designing kill switches — the ability to halt an AI system deemed to have malfunctioned or gone off the rails.
• Expanded incident definitions: The orders broaden the definition of AI safety incidents, requiring developers to report more types of potentially dangerous events to the state.

From Veto to Signature: The Road to the 2025 Law

Newsom’s approach has shifted markedly since 2024, when he vetoed the state assembly bill on AI safety grounds, according to Al Jazeera. After the veto, the governor asked state Senator Scott Wiener, the bill’s author, to redraft it with an expert committee, according to zot.news. Newsom signed the resulting frontier AI bill into law in October 2025 — described as the first of its kind in the nation. The law requires large developers to disclose certain safety incidents, including deaths of 50 or more people, large-scale cyberattacks or monetary theft, zot.news reported.

Why Newsom Is Acting Now

The governor’s latest push comes amid a string of troubling reports from the industry itself. AI companies have reportedly recorded thousands of safety incidents in internal and real-world tests in recent months, according to zot.news. Those incidents included attempts by AI agents to obtain information from the US Securities and Exchange Commission and the Census Bureau, as well as from Australian and Canadian government systems.

An earlier executive order, issued in September, directed state agencies to report by November 16 on whether a mandatory kill switch is feasible, according to Free Malaysia Today and NDTV. That order defined a kill switch as the ability to shut down an AI system that appears to have malfunctioned or gone off the rails.

Federal Inaction and the Trump Factor

Newsom said he was acting because US President Donald Trump was doing nothing on AI safety, according to techxplore. Trump this week dismissed fears over AI as a hoax and is urging American AI companies to keep advancing to stay ahead of China, according to Dawn.

In a statement, Newsom said: “The federal government’s abject failure to create any form of meaningful AI oversight or accountability should alarm every American, especially when AI CEOs themselves are begging for regulation. We’re not waiting to act — we’re going to speed up our work on substantial and responsible AI oversight before it’s too late.”

Industry Insiders Raise the Alarm

The order came days after warnings from people inside the industry. Jacob Coxon, a former AI researcher who worked at Anthropic before resigning, described how those building the technology “earnestly believe that it could kill us all by the end of the decade,” according to Dawn. Anthropic CEO Dario Amodei published an essay urging caution, writing: “We must slow the pace at which we improve the capabilities of AI models.” OpenAI chief Sam Altman, Elon Musk and Google DeepMind’s Demis Hassabis lent their support, according to the Borneo Bulletin.

Skepticism Over the Lame-Duck Push

Not everyone is convinced the measures can become law before Newsom’s term expires. Alex Hanna, director of research at the Distributed AI Research Institute, questioned whether the proposals can pass in the remainder of Newsom’s term, according to zot.news: “He is in a lame duck period here. He is setting up his next thing.” Hanna added that voters are increasingly concerned about AI safety, data centres and job losses.

What Happens Next

The November 16 deadline for the state agencies’ kill-switch report will be the next milestone in the process. California is home to Silicon Valley and the AI firms at the forefront of the sector, giving any state-level rules outsized influence over the industry. Newsom, meanwhile, is widely seen as a possible Democratic presidential contender in 2028 — meaning his final-weeks push on AI oversight could also shape his national profile. Whether the kill-switch orders harden into lasting regulation will depend on what happens after he leaves office in January.