California Governor Vetoes AI Safety Bill

By David Shepardson and Anna Tong
WASHINGTON/SAN FRANCISCO (Reuters)

California Governor Gavin Newsom vetoed a contentious artificial intelligence safety bill, citing concerns from the tech industry that it could hinder innovation and drive AI companies out of the state.

Newsom remarked that the proposed legislation “does not take into account whether an AI system is deployed in high-risk environments, involves critical decision-making, or the use of sensitive data.” He criticized the bill’s stringent standards, which he argued would apply even to basic AI functions.

After consulting experts on generative AI, Newsom emphasized the need for empirical, science-based risk assessments instead of the proposed rigid regulatory framework. He ordered state agencies to expand evaluations of potential catastrophic risks associated with AI usage.

Generative AI has generated both excitement and fears regarding its implications for job markets, elections, and social dynamics.

Senator Scott Wiener, who authored the bill, stated that without such regulations, public safety is at risk as powerful technologies evolve. He criticized the veto, asserting that it leaves no binding rules for companies developing advanced AI technologies.

Newsom expressed the urgency of proactive safety measures but disagreed with adopting a solution lacking empirical analysis. He plans to collaborate with the legislature on future AI regulations, especially in light of stalled federal legislation.

The tech industry coalition, Chamber of Progress, commended the veto, arguing that California’s tech sector thrives on competition and transparency.

The rejected bill would have mandated safety testing for advanced AI models and included a “kill switch” mechanism for controlling AI operations. It aimed to create a state body to oversee the development of advanced AI systems known as “Frontier Models.”

The proposal faced opposition from major tech firms like Google, OpenAI, and Meta, alongside some Congressional Democrats. In contrast, proponents included figures like Elon Musk and companies like Anthropic, which expressed conditional support for the bill.

Additionally, Newsom signed legislation to assess Generative AI threats to California’s critical infrastructure, beginning with examinations of energy and water sectors before moving to communication providers in the upcoming year.