The California State Assembly has approved the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act (SB 1047), marking a significant step toward AI regulation in the United States.
The bill requires artificial intelligence companies in the state to implement safety measures before training advanced models. These include the ability to quickly shut down models, protect against harmful modifications, and maintain testing procedures to assess risks.
Senator Scott Wiener, the bill’s author, emphasised that SB 1047 is designed to ensure AI labs conduct necessary safety checks.
The bill has faced criticism, particularly from OpenAI, Anthropic, and others, who argue that it could disproportionately affect small, open-source developers.
In response, the bill was amended to replace criminal penalties with civil ones and to refine enforcement provisions.
The State Senate is expected to pass the amended bill, which will then go to Governor Gavin Newsom for approval by the end of September.
Elon Musk, a co-founder and early backer of OpenAI, earlier this week endorsed SB 1047.
On his platform X (formerly Twitter), he said: "This is a tough call and will make some people upset, but, all things considered, I think California should probably pass the SB 1047 AI safety bill.
"For over 20 years, I have been an advocate for AI regulation, just as we regulate any product/technology that is a potential risk to the public."
Musk's antipathy towards OpenAI (and to some extent his fears around artificial intelligence) are captured in his recent lawsuits.