California's debate over AI regulation has reached a decisive moment. Governor Gavin Newsom said Tuesday that 38 artificial intelligence bills are on his desk, but SB 1047 has become the measure drawing the sharpest attention from industry, lawmakers and AI safety advocates.
The bill is designed to reduce the risk that powerful AI systems could contribute to catastrophic harm. Newsom has not said whether he will sign or veto it, but his public comments made clear that he sees difficult tradeoffs in the proposal.
A bill aimed at extreme AI risks
SB 1047 focuses on high-consequence dangers. According to the source article, the bill tries to prevent AI systems from causing catastrophes and would hold large AI vendors liable if their products are used to cause grievous harm, including damage such as bringing down critical infrastructure.
The measure has become controversial because it is not primarily centered on the AI problems already visible in daily life. Its critics argue that it targets events such as mass casualty incidents and cybersecurity events costing more than $500 million, while doing little to address conduct that falls short of that threshold.
That distinction matters for the policy debate. A law built around catastrophic risk can signal that the state is taking future dangers seriously. But it can also raise questions about whether regulators are spending their effort on the right problems, especially when AI-generated election misinformation and AI clones of actors are already part of the public record described in the source article.
Newsom is balancing risk and industry growth
Newsom discussed the issue Tuesday with Salesforce CEO Marc Benioff onstage at the 2024 Dreamforce conference in San Francisco. His remarks framed the challenge as a search for regulation that controls real dangers without punishing productive experimentation.
“We’ve been working over the last couple years to come up with some rational regulation that supports risk-taking, but not recklessness,”
He said that task is especially difficult with SB 1047 because of the possible effect on the AI sector and the open source community. The governor described the bill as having a potentially outsized impact and referred to a possible chilling effect.
Newsom also emphasized a distinction between risks that can already be demonstrated and risks that remain hypothetical. Later, he added, “I can’t solve for everything. What can we solve for?”
Those comments do not amount to a veto threat. But they show that Newsom is not treating AI safety as a simple question of being for or against regulation. He is asking whether this specific bill fits the kind of near-term, evidence-based lawmaking he appears to prefer.
Other AI bills show where the governor has acted
The source article notes that Newsom signed five AI-related bills earlier on Tuesday. Those laws address problems that have already appeared in 2024, including AI-generated election misinformation and Hollywood studios creating AI clones of actors.
That contrast helps explain the political pressure around SB 1047. The governor has already shown willingness to sign AI rules. The unresolved question is whether he sees catastrophe-focused liability as the right next step.
For Newsom, California's position in the AI industry is part of the calculation. He said the federal government has “failed to regulate” in the AI space, and he pointed to California's earlier role in technology regulation, including social media and privacy.
“[AI] is a space where we dominate, and I want to maintain our dominance,”
At the same time, he said he feels responsibility to address extreme concerns that even strong supporters of the technology have raised. That leaves him trying to preserve California's lead while responding to fears about where AI systems could go wrong.
The political split around SB 1047
The coalition around SB 1047 is unusual because support and opposition do not fall neatly into a single camp. OpenAI, Nancy Pelosi, the United States Chamber of Commerce and Big Tech trade groups are pushing Newsom to veto the bill.
On the other side, Elon Musk and Anthropic have expressed tepid enthusiasm. Some AI researchers, including Yoshua Bengio and Geoffrey Hinton, have fully endorsed SB 1047.
State senator Scott Wiener, who supports the bill, framed the issue as one of California leadership when federal action is absent.
“Governor Newsom understands better than anyone the importance of California’s leadership when the federal government does not step up,”
Wiener also said he had confidence that Newsom would give the bill the consideration it deserves. Nathan Calvin, senior policy counsel for the Center for AI Safety Action Fund, also expressed hope that Newsom would sign it.
“We remain hopeful that the Governor will sign SB 1047 because he understands the bottom line - if California won’t lead on safe and responsible AI innovation, who will?,”
What the decision will signal
Newsom told the LA Times he has yet to make up his mind on SB 1047. The source article says he has two weeks to decide.
Whatever he chooses, the decision will send a message about what kind of AI regulation California wants to prioritize. Signing SB 1047 would show that the state is willing to legislate around severe potential harms before they happen. Vetoing it would suggest that Newsom wants AI rules tied more closely to harms already visible today.
The governor also suggested that the immediate impact of signing SB 1047 may have been overstated. But he warned that choosing the wrong bills over several years could deeply affect California's AI dominance.
That is the core tension. California is being asked to lead because federal regulation has not arrived, but the state is also home to an AI industry that Newsom says he wants to keep strong. SB 1047 sits directly at that intersection: safety, liability, innovation and political judgment, all compressed into one pending decision.