California Gov. Gavin Newsom on Sunday vetoed a landmark bill aimed at establishing the nation’s first safety measures for large artificial intelligence models.
The decision is a major blow to efforts to control a domestic industry that is developing rapidly with little oversight. bill It may have established some of the first regulations. Supporters said it “researched large-scale AI models in the country and paved the way for AI safety regulations across the country.”
Earlier this month, the Democratic governor told an audience at Dreamforce, an annual conference hosted by software giant Salesforce, that California must take the lead on AI regulation in the face of federal inaction, but that the proposal would have a “chilling effect on the industry.” “It can be crazy,” he said.
Newsom said the proposal, which has drawn fierce opposition from startups, tech giants and some House Democrats, could harm domestic industries by setting stringent requirements.
“SB 1047, while well-intentioned, does not consider whether AI systems are deployed in high-risk environments, require critical decisions, or involve the use of sensitive data,” Newsom said in a statement. “Instead, this bill imposes strict standards for even the most basic functions as long as large systems deploy them. I do not believe this is the best approach to protecting the public from real threats posed by technology.”
Instead, Newsom announced Sunday that the state will work with several industry experts, including AI pioneer Fei-Fei Li, to develop guardrails for powerful AI models. Lee opposed the AI safety proposal.
The bill, aimed at reducing potential risks posed by AI, would have required companies to test their models and disclose safety protocols to prevent them from being manipulated to destroy the national power grid or create chemical weapons. Experts say this scenario could be possible in the future as the industry continues to evolve rapidly. It would also have provided workers with protections for whistleblowers.
The bill is one of a number of bills passed by Congress this year to regulate AI. deepfake And protect workers. State lawmakers said action must be taken this year, citing difficult lessons California learned from failing to rein in social media companies when it had the chance.
Supporters of the bill, including Elon Musk and Anthropic, say the proposal could have instilled some level of transparency and accountability for large-scale AI models, as developers and experts still do not fully understand how they work and how they operate. I said there is. why.
The bill targeted systems that required more than $100 million to build. Current AI models have never reached this benchmark, but some experts say that could change within the next year.
“This is due to the massive expansion of investment within the industry,” said Daniel Kokotajlo, a former OpenAI researcher who resigned in April after accusing the company of ignoring AI risks. “This is an enormous amount of power for private companies to control irresponsibly, and it is also incredibly dangerous.”
The United States already lags behind Europe in regulating AI to limit risks. California’s proposal, while not as comprehensive as Europe’s regulations, would have been a good first step in establishing shields for a fast-growing technology that has raised concerns about job losses, misinformation, privacy violations and automation bias, supporters said.
Last year, a number of leading AI companies voluntarily agreed to follow safeguards set by the White House, including testing and sharing information about their models. The California bill would have mandated that AI developers follow those promises and similar requirements, supporters of the bill said.
But critics, including former House Speaker Nancy Pelosi, argued the bill would “kill California technology” and stifle innovation. It would have prevented AI developers from investing in large-scale models or sharing open source software, they said.
Newsom’s decision to veto the bill marks another victory for big tech companies and AI developers in California. Many of them, along with the California Chamber of Commerce, have lobbied the governor and lawmakers over the past year to block them from advancing AI regulations.
Two other comprehensive AI proposals that faced fierce opposition from the tech industry and others were killed last month ahead of a legislative deadline. The bill would have required AI developers to label AI-generated content and prohibit discrimination on AI tools used to make hiring decisions.
