California’s Newsom signs executive order to explore new AI rules, consider ‘kill switch’
California Gov. Gavin Newsom issued an executive order on Friday to speed up two AI laws he recently signed and explore proposals like requiring that AI companies develop a “kill switch” for their most advanced models.
The order comes after the Democrat told POLITICO this week he was mulling his options to take further action on AI safety, including a possible special legislative session to address recent safety concerns set off by AI researchers’ warnings that the technology could pose an existential risk to humanity and prominent autonomous hacking incidents.
Newsom said his executive action was also driven by federal inaction, after President Donald Trump responded to the warnings by dismissing them as a “hoax” and saying the only guardrails AI needs are his presidency. On Friday, Newsom urged Trump and Congress to adopt California’s AI framework or use the state’s existing regulations as a floor for federal rules.
“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,” Newsom said in his press release.
The order moves up the implementation of two California laws that Newsom signed this month, SB 813 and AB 1405, regulating how outside groups evaluate AI programs for safety.
It also convenes experts to provide California with recommendations within two months on how the state could strengthen existing AI safety and security laws.
Among the ideas they would consider is updating bill definitions of critical safety events to include incidents where models lose control, such as one where OpenAI’s model hacked into Hugging Face.
Another is the kill switch mandate for frontier AI models, a version of which has been introduced in Congress but blocked on the Senate floor. The idea also garnered support from LinkedIn co-founder and Democratic megadonor Reid Hoffman this week at a POLITICO event. Some others in the tech industry still oppose kill switches, arguing they are not technically feasible.
Anthropic CEO Dario Amodei, who called for a slowdown in AI development amid existential risk concerns, said Monday that while a kill switch “could be a good idea,” it was no “panacea.”
In 2024, Newsom vetoed a bill that proposed pre-release safety testing and a kill switch for certain large AI models in the event that the systems were misused or went rogue.
After the veto, Newsom instead tasked an expert panel with recommending a solution. The resulting law, SB 53, authored by state Sen. Scott Wiener and approved by Newsom last year, positioned California as a national leader without antagonizing as badly the Silicon Valley interests that rallied against the vetoed measure.
AI safety advocates and some state legislators have, however, warned SB 53 would not force the disclosure of autonomous hacks like the Hugging Face incident.
