California Governor Gavin Newsom has vetoed a controversial attempt to regulate artificial intelligence (AI), expressing concerns that the bill could hinder innovation. The decision came after intense pressure from tech companies and was announced at the last minute, following the bill’s passage through the state legislature in August.
The proposed legislation would have required developers of powerful AI models to adhere to strict rules, including implementing a kill switch as a safety measure. However, leading AI companies such as Google, OpenAI, and Meta strongly opposed the bill and actively lobbied against it. They argued that premature regulation could impede AI development and jeopardize California’s prominent role in advancing this technology. On the other hand, Amazon-backed Anthropic and Elon Musk’s start-up xAI supported the legislation.
In a letter addressed to the state senate on Sunday, Governor Newsom defended his veto of SB 1047 (also known as Safe and Secure Innovation for Frontier Artificial Intelligence Systems Act). He stated that such regulations could limit innovation for public benefit while highlighting California’s position as home to 32 leading AI companies worldwide.
Governor Newsom specifically criticized targeting models based on their size—models costing over $100 million would have been subject to safety testing and other safeguards—as an inadequate metric. He argued that this approach might create a false sense of security since smaller specialized models could be equally or even more dangerous.
Senator Scott Wiener, who introduced the bill, expressed disappointment with Newsom’s decision. Wiener believed that oversight of massive corporations making critical decisions affecting public safety should be prioritized but acknowledged Congress’ lack of meaningful regulation in this area.
Newsom countered by stating that SB 1047 did not adequately consider factors such as high-risk environments or critical decision-making when deploying an AI system or handling sensitive data. He emphasized that stringent standards applied even to basic functions if used by large systems—an approach he deemed ineffective for protecting against real threats posed by AI technology.
In recent weeks, Governor Newsom has signed bills related to generative AI technology deployment and regulation concerning deepfakes, AI watermarking, and misinformation. Additionally, experts have collaborated with California authorities in developing practical guidelines for deploying generative AI backed by empirical evidence.
The governor’s veto drew criticism from think-tank Artificial Intelligence Policy Institute (AIPI). Executive director Daniel Colson described it as misguided and reckless while urging lawmakers to return next session with sensible measures ensuring responsible development of AI technologies.
