California is on the verge of passing a historic artificial intelligence regulation that could set the tone for AI safety measures across the United States. Senate Bill 1047, the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act, which mandates testing and monitoring of AI systems costing over $100 million to develop, was passed by the California Assembly on August 28. The state Senate had already approved the bill in May.

The bill now heads to Governor Gavin Newsom, who must decide by the end of September whether to sign it into law, veto it, or allow it to become law without his signature. If passed, it would be the first legislation in the US to impose safety requirements on large-scale AI systems. California, home to 35 of the world’s top 50 AI companies—including tech giants like Google, Apple, Meta, and OpenAI—would become a global leader in AI regulation.

Proposed by State Senator Scott Wiener, a Democrat representing San Francisco, SB 1047 has met resistance from major tech firms, including OpenAI, Google, and Meta. Critics argue the legislation could hinder innovation in a rapidly evolving industry. Several state politicians, including representatives Anna Eshoo and Zoe Lofgren, have urged Newsom to veto the bill, citing concerns over its potential impact on California’s tech economy.

Despite opposition, the bill has garnered support from AI researchers and employees who have expressed concerns about the unchecked development of powerful AI systems. SB 1047 includes whistleblower protections for employees who report issues with AI systems, addressing a key demand made in an open letter signed by prominent AI figures such as Yoshua Bengio, Geoffrey Hinton, and Stuart Russell.

Meanwhile, OpenAI and Anthropic recently became the first AI companies to partner with the US government’s Intelligence Safety Institute, allowing their models to undergo pre-release testing for safety. This development comes as Newsom contemplates whether California should take the lead in AI regulation or defer to federal efforts, with OpenAI’s CEO Sam Altman advocating for national-level oversight.

Newsom has not yet indicated whether he will sign SB 1047, but the decision will likely have far-reaching implications for both the tech industry and the future of AI governance in the United States.