California Governor Vetoes Controversial AI Safety Bill, According to Reuters

By David Shepardson and Anna Tong

WASHINGTON/SAN FRANCISCO – California Governor Gavin Newsom vetoed a controversial artificial intelligence safety bill on Sunday, responding to concerns from the tech industry that it could drive AI companies out of the state and impede innovation.

In his statement, Newsom expressed that the legislation did not adequately consider the deployment of AI systems in high-risk environments, critical decision-making processes, or handling of sensitive data. He noted that the bill would impose "stringent standards" even on basic AI functions used by large systems.

The Governor has engaged leading experts on generative AI to assist California in creating effective guidelines that focus on empirical, science-based analysis of technological trajectories. He also directed state agencies to broaden their evaluations of risks associated with potential catastrophic events linked to AI usage.

The rise of generative AI—which is capable of producing text, images, and videos in response to prompts—has generated both excitement and concern, paving the way for debates regarding job displacement, election impacts, and more serious threats.

State Senator Scott Wiener, the bill’s author, argued that the legislation was essential to protect the public from the unchecked growth of AI technologies. He stated that the veto diminishes safety and allows companies working with powerful technologies to operate without strict regulations. Wiener criticized the reliance on voluntary industry commitments, highlighting their frequent ineffectiveness in safeguarding public interests.

While Newsom acknowledged the urgency of action to protect public safety, he rejected the notion that a solution should be adopted without comprehensive analysis of AI capabilities.

He indicated that he plans to collaborate with the legislature to address AI regulations in the upcoming session. His comments come amid stalled efforts in Congress to establish regulatory safeguards, while the Biden administration pushes forward with proposals for AI oversight.

In his statement, Newsom suggested that a state-specific approach may be necessary, particularly in the absence of federal action.

The Chamber of Progress, a coalition representing the tech industry, commended Newsom’s decision, reaffirming that California’s tech economy relies on competition and openness.

The rejected bill would have required rigorous safety testing for advanced AI models, particularly those exceeding $100 million in development costs or possessing significant computing requirements. It also called for developers to implement fail-safe mechanisms for deactivating AI models and proposed the creation of a state body to oversee the development of next-generation "Frontier Models."

The legislation faced significant opposition from various stakeholders, including major tech companies like Google, OpenAI, and Meta, who expressed concerns about its implications. Some lawmakers, including Representative Nancy Pelosi, also opposed the initiative, while supporters included prominent figures like Elon Musk, who highlighted the potential benefits of the bill.

In a related move, Newsom signed new legislation requiring an assessment of the risks posed by generative AI to California’s critical infrastructure. The state is currently evaluating risks related to energy infrastructure and plans to extend these assessments to water and communication sectors in the coming year.