California Governor Gavin Newsom vetoed the controversial artificial intelligence bill, saying that being too innovative would create uncontrollable risks

California Governor Gavin Newsom has vetoed the highly debated Safe and Secure Innovation for Frontier Artificial Intelligence Models Act (SB 1047), which sought to implement stringent regulations on artificial intelligence (AI) development. On September 30, Newsom rejected the bill, arguing that while AI safety is critical, the proposed law would stifle innovation and fail to address the real dangers posed by the technology.

The bill, spearheaded by San Francisco Senator Scott Wiener, would have required mandatory safety testing for AI models developed by major companies like OpenAI, Meta, and Google. It also included provisions for a “kill switch” to deactivate AI systems in case of extreme risk and would have made AI developers legally liable if their models posed ongoing threats, such as hypothetical AI grid takeovers.

Pushback from Silicon Valley

The bill had faced strong opposition from Silicon Valley and tech industry leaders, who feared it would burden AI development with excessive regulations. Critics, including major tech firms and AI policymakers, argued that the legislation could significantly hamper the sector’s growth in California.

Newsom echoed these concerns in his statement, saying the bill focused too much on regulating top AI firms and basic functionalities without addressing the more significant risks that come with AI.
He explained, “I do not believe this is the best approach to protecting the public from real threats posed by the technology.”

Balancing Innovation and Safety

Newsom stressed that while he rejected SB 1047, he still acknowledged the need for AI safety measures. He said he had enlisted the help of top AI safety experts to guide California in creating “workable guardrails” that would be informed by science-based risk assessments.

Newsom added that state agencies were tasked with expanding their evaluation of AI’s potential catastrophic risks to ensure proper safeguards are implemented. He emphasized that regulators must not wait for a disaster before taking action to protect the public.
Despite his veto, Newsom’s administration has been active on AI regulation, having signed more than 18 AI-related bills in the last 30 days.

Related reading：The Cognitive Revolution: Summary of "When AI Will Outthink Humans"

Divided Reactions

The veto sparked mixed reactions. Prominent figures, such as House Speaker Nancy Pelosi and companies like OpenAI, had voiced opposition to the bill, warning it could hinder AI innovation.

On the other hand, Elon Musk, who is developing his own AI model, Grok, expressed cautious support for the bill, advocating for broader AI regulations. Musk acknowledged that passing SB 1047 would be a “tough call” but indicated he leaned toward supporting stronger safety measures.
The bill’s broad implications—targeting models costing over $100 million—raised concerns that it could later expand to cover smaller AI developers, potentially stifling innovation across the board.
While Newsom’s veto may halt the specific provisions of SB 1047, the conversation around AI regulation is far from over, as California looks for more balanced approaches to safeguard the public while allowing the industry to thrive.