SB 1047, the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act, was a California bill authored by State Senator Scott Wiener that would have imposed safety obligations on the developers of the largest AI models and on the companies providing the computing power to train them. On September 29, 2024, Governor Gavin Newsom returned the bill to the State Senate without his signature. The veto ended what had been the most closely watched state-level AI safety effort in the United States, after a campaign that split Silicon Valley between supporters who saw it as overdue and opponents who argued it would chill innovation.
In his veto message, Newsom wrote: “I am returning Senate Bill 1047 without my signature.” He described what the bill would have done - requiring developers of large AI models “to put certain safeguards and policies in place to prevent catastrophic harm” and establishing “the Board of Frontier Models - a state entity - to oversee the development of these models.” He noted that “California is home to 32 of the world’s 50 leading AI companies.”
His central objection was the bill’s threshold for regulation. Newsom argued the bill focused “only on the most expensive and large-scale models” and could “give the public a false sense of security about controlling this fast-moving technology,” because “smaller, specialized models may emerge as equally or even more dangerous.” He wrote that SB 1047 “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,” and instead “applies stringent standards to even the most basic functions.” He added: “Let me be clear - I agree with the author - we cannot afford to wait for a major catastrophe to occur before taking action to protect the public,” but said any framework “needs to keep pace with the technology itself.”
Why business readers should care: the veto framed a debate that recurs across AI governance - whether to regulate by the size and cost of a model or by the risk of how it is used. The bill’s compute-and-cost trigger, rather than a use-based or harm-based one, was the explicit reason the governor gave for rejecting it. One year later the same governor signed a narrower transparency-focused successor, 2025-sb-53-frontier-ai-act, illustrating how the regulatory line in the United States shifted toward disclosure rather than pre-deployment safety mandates.