Newsom Proposes Sweeping Ai Rules As ‘template’ For Washington
Gavin Newsom unveiled a sweeping proposal Wednesday to give California broad powers to regulate the AI industry and impose strict safety measures, casting it as a model for the nation as he prepares for a likely presidential run.
The plan — which would allow California officials to intervene in a company’s work if they determine AI models are at risk of going rogue, embed independent safety monitors inside labs and pursue “emergency brakes” to shut down dangerous AI models — is being pitched as a roadmap to legislation he wants state lawmakers to pass.
“This is an action document. I’m already working with members of our Legislature to introduce the legislative frameworks to move quickly and make real what we're promoting and promising in this report,” Newsom told reporters and members of a closed-door roundtable the governor convened in Washington.
The proposal is the latest move to rein in the explosive technology by a governor who has taken a populist turn on tech in recent months, signing a suite ofdata center bills, issuing an executive order to explore “kill switch” requirements and proposing anAI sovereign wealth fund. It’s another step in what has been a recalibration for a governor who has touted the industry as a source of jobs and revenue, and who has longtime allies in Silicon Valley — but who saw public opinion turn sharply against AI in recent months, culminating with its drubbing in the midterms.
“This takes things to the next level, not only of conversation, but of consciousness in terms of how we effectively regulate artificial intelligence, not only in my home state, but we believe a template for how we should be doing it federally as well,” Newsom said.
If Newsom’s report represents a final effort to shape AI policy in California before leaving office in January, it also served as the latest signal of his likely presidential run. He released the proposal in downtown Washington — not Sacramento or San Francisco — in front of a swarm of national political and tech reporters.
It also offered him a chance to cast California as a counterweight to President Donald Trump’s resistance to strict AI regulation and shape national debates on the issue ahead of his likely 2028 presidential bid.
“This is not a Truth Social or a tweet,” Newsom said, casting the action as necessary “in the absence of any federal leadership.”
Newsom’s report, “AI Safety in 2026 and Beyond,” capped a fast-track inquiry he launched in September when he issued the executive order directing state agencies and a panel of experts to present him with recommendations on kill switches and other AI-safety issues.
The four experts Newsom appointed — Stanford professor Rob Reich, former Obama White House Chief Digital Officer Jason Goldman, economist Gillian Hadfield and Alondra Nelson, who was Office of Science and Technology Policy director under former President Joe Biden — were given a Monday deadline to send the governor their suggested path forward. Hadfield joined Newsom for the roundtable on the proposal Wednesday, along with Nelson, the governor’s business and economic development head Dee Dee Myers, former Office of Management and Budget Director Shalanda Young, and other officials and academics.
Newsom’s report inevitably will also raise questions about his AI record and whether he can effectively pivot from his past with a proposal that grapples with ideas he shot down in 2024, when he vetoed a major AI safety bill authored by Democratic Rep.-elect Scott Wiener, then a state senator. The governor has previously said he does not regret vetoing the legislation, citing unintended consequences.
The report handed down Wednesday outlines plans for expanding California’s current frontier AI safety and transparency law, SB 53, and addresses several of the most pressing AI policy questions being debated in Washington and Sacramento.
It calls for giving the state the authority to cap companies’ ability to use automation in developing advanced AI, a process known as recursive self-improvement. Under the governor’s plan, companies also would be required to work closely with independent auditors to regularly run safety tests and create a plan for shutting down dangerous models in emergency situations. And if developers don’t act quickly, the report says state officials should be able to order immediate shutdowns.
The report stopped short of calling for the development of so-called kill switches, saying the idea of a “big red button” used to shut down a rogue AI model does not capture a broader array of controls California must consider. Those additional controls include limiting users’ access to models, coordinating shutdowns of the servers they run on and suspending development of certain models.
Major AI developers and safety advocates are closely tracking Newsom’s moves. OpenAI, Anthropic and Meta sent the governor’s office detailed policy proposals in an effort to shape his thinking, as did AI-focused nonprofits like Fathom and auditing firms that assess risks posed by frontier AI models.
Experts’ recommendations offer the California Legislature and Newsom’s successor, Gov.-elect Xavier Becerra, a roadmap for building out SB 53, but their ideas are not guaranteed to become law. The state Assembly and state Senate held separate hearings last month to help inform their own approaches to strengthening AI regulations and may choose to ignore Newsom’s last-minute recommendations if they wish. Though Newsom told reporters he was working with lawmakers to advance the proposals, his final term ends in less than two months.
Popular Products
-
Electric Toothbrush & Water Flosser S...$43.56$21.78 -
Foldable Car Trunk Multi-Compartment ...$329.56$164.78 -
Mommy Diaper Backpack with Stroller O...$111.56$55.78 -
Ai Dash Cam with Front & Rear HD, GPS...$295.56$147.78 -
Smart Auto-Recharge Robot Vacuum Cleaner$613.56$306.78