California Governor Gavin Newsom has signed SB 53, the country’s first legislation to set new transparency requirements for large AI companies.
SB 53, which passed the state legislature two weeks ago, requires large AI labs, including Openai, Humanity, Meta and Google Deepmind, to be transparent about safety protocols. It also guarantees the protection of whistleblowers for employees of these companies.
Additionally, SB 53 creates a mechanism for AI companies and the public to report potentially serious safety incidents to the California Department of Emergency Services. In addition, businesses must report incidents related to crimes committed without human supervision, such as cyber attacks, based on models not required by EU AI law.
The bill has undergone mixed reactions from the AI industry. While tech companies widely argue that state-level AI policies risk creating a “regulatory patchwork” that hinder innovation, humanity has supported the bill. Meta and Openai lobbyed against it. Openai wrote an open letter to Gov. Newsom, which discouraged the signature of SB 53.
The new bill comes as some of Silicon Valley’s tech elites have poured hundreds of millions of people into super PACs into back candidates who support their light-up approach to AI regulation. In recent weeks, Openai and Meta leaders have launched a pro-ai super PACS aimed at supporting AI-friendly candidates and bills.
Still, other states may be seeking inspiration from California to seek to curb the potential harm caused by unmitigated advances in such powerful emerging technologies. In New York, a similar bill has been passed by state lawmakers awaiting Gov. Kathy Hochul’s signature or veto.
“California is proving that we can establish regulations to protect our communities and ensure that the growing AI industry continues to thrive,” Newsom said in a statement. “This law balances that. AI is a new frontier of innovation, and California is not only here, but also strong as a national leader by enacting the national frontier AI safety law that builds public trust as this new technology rapidly evolves.”
TechCrunch Events
San Francisco
|
October 27th-29th, 2025
The governor is also weighing another bill (SB 243) this month with bipartisan support from both the state legislature and the Senate. The bill regulates AI companion chatbots, requires operators to implement safety protocols, and requires that bots be legally accountable if they do not meet those standards.
SB 53 is Sen. Scott Winner’s second attempt on the AI Safety Bill after Newsom rejected his more drastic SB 1047 last year amid a major pushback from AI companies. With this bill, Wiener reached out to major AI companies to understand the changes they made to the bill.
