AUS
NZ
UK
CA
USA
IND
The California State Assembly has given the green light to the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act (SB 1047), a pivotal move towards regulating artificial intelligence within the United States.
This legislation mandates that AI companies operating in California implement robust safety protocols before advancing their model training processes. Key requirements include the ability to swiftly deactivate models, safeguard against harmful alterations, and consistently evaluate risks through ongoing testing.
Senator Scott Wiener, who authored the bill, highlighted that SB 1047 aims to ensure that AI laboratories perform essential safety checks, maintaining a secure development environment.
However, the bill has encountered criticism, particularly from major players like OpenAI, Anthropic, and others, who argue that the regulations could disproportionately impact smaller, open-source developers. In response to these concerns, the bill was revised, substituting criminal penalties with civil ones and fine-tuning the enforcement mechanisms.
The State Senate is anticipated to pass the revised bill, which will then be presented to Governor Gavin Newsom for approval by the end of September.
Elon Musk, co-founder and early supporter of OpenAI, voiced his support for SB 1047 earlier this week. Speaking on his platform X (formerly known as Twitter), he remarked, "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."
Musk, who has long advocated for AI regulation, drew parallels between AI and other potentially risky technologies that are subject to regulation. His endorsement reflects a broader concern about the potential dangers of AI, a sentiment he has expressed through recent legal actions.
Musk's relationship with OpenAI has been complex, marked by both his early involvement and ongoing concerns about the future of artificial intelligence. His support for SB 1047 underscores a cautious approach to AI development, one that prioritizes public safety and regulatory oversight.
