California Governor Gavin Newsom vetoed a contentious artificial intelligence safety bill on Sunday, following tech industry objections that it could drive AI companies out of the state and impede innovation. Newsom called for leading Generative AI experts to assist California in "developing workable guardrails" with a focus on "empirical, science-based trajectory analysis." Additionally, he directed state agencies to broaden their assessment of risks associated with potential catastrophic AI events.

Generative AI, capable of producing text, photos, and videos from open-ended prompts, has raised both excitement and concerns about job displacement, election interference, and the possibility of overpowering human control with catastrophic consequences. Democratic State Senator Scott Wiener, the bill's author, argued that legislation was essential to protect the public before AI advancements become unwieldy or uncontrollable.

The AI industry is rapidly growing in California, and some leaders have questioned the bill's impact on these companies. Wiener stated that the veto makes California less safe and implies that "companies creating extremely powerful technology face no binding restrictions." He criticized voluntary industry commitments as unenforceable and often ineffective for public protection.

Newsom emphasized the importance of taking action before a major catastrophe occurs but disagreed with settling for a solution not grounded in empirical AI analysis. He pledged to collaborate with the legislature on AI legislation in the upcoming session. This comes amid stalled U.S. Congress legislation for safeguards and the Biden administration's push for regulatory AI oversight.

Newsom suggested that a California-specific approach might be necessary in the absence of federal action by Congress. The tech industry coalition, Chamber of Progress, praised Newsom's veto, stating that "California's tech economy has always thrived on competition and openness."

The bill would have required safety testing for advanced AI models costing over $100 million or requiring significant computing power. Developers would also need to provide methods for disabling AI models, effectively a kill switch. It proposed the establishment of a state entity to oversee the development of "Frontier Models" exceeding current capabilities.

The bill faced opposition from Alphabet's Google, Microsoft-backed OpenAI, and Meta Platforms, all developing generative AI models. Some Democrats in U.S. Congress, including Representative Nancy Pelosi, also opposed the bill. Proponents included Tesla CEO Elon Musk, who runs an AI firm called xAI. Amazon-backed Anthropic acknowledged the bill's benefits but raised concerns about certain aspects.

Newsom also signed legislation requiring the state to assess potential threats from Generative AI to California's critical infrastructure. The state is analyzing energy infrastructure risks and will extend the same risk assessment to water and communications infrastructure in the coming years.