Sacramento, California – Governor Gavin Newsom has signed SB 53 into law, establishing California as the first U.S. state to implement broad transparency and safety requirements for large artificial intelligence (AI) companies. The legislation mandates specific protocols for prominent AI developers, including OpenAI, Anthropic, Meta, and Google DeepMind, and introduces whistleblower protections for employees within these firms.
SB 53, which gained legislative approval two weeks prior, outlines several key provisions. It requires AI laboratories to disclose safety protocols and creates a reporting mechanism for potential critical safety incidents to California’s Office of Emergency Services. Companies are also obligated to report incidents involving crimes committed without human oversight, such as cyberattacks, and instances of deceptive behavior by an AI model that are not covered under the existing EU AI Act.
The bill has elicited varied reactions across the AI industry. Some technology firms have expressed concerns that state-level AI policies could lead to a fragmented regulatory landscape, potentially impeding innovation. OpenAI and Meta actively lobbied against the bill, with OpenAI publishing an open letter to Governor Newsom urging a veto. Conversely, Anthropic publicly endorsed the legislation.
The legislative action unfolds amidst significant financial contributions from some Silicon Valley leaders, who have invested millions in super PACs supporting candidates and policies favorable to a less restrictive approach to AI regulation. Leaders at OpenAI and Meta have launched pro-AI super PACs in recent weeks, aiming to back candidates and bills that align with their industry perspectives.
Despite industry pushback, the passage of SB 53 is anticipated to influence other states considering similar regulatory frameworks for emerging AI technologies. A comparable bill in New York has passed state lawmakers and awaits action from Governor Kathy Hochul. Governor Newsom stated that the legislation "strikes that balance" between protecting communities and ensuring the AI industry's growth, positioning California as a "national leader by enacting the first-in-the-nation frontier AI safety legislation."
Separately, Governor Newsom is evaluating SB 243, a bill passed with bipartisan support that seeks to regulate AI companion chatbots. This proposed legislation would require operators to implement safety protocols and could hold them legally accountable for failures to meet these standards. SB 53 marks Senator Scott Wiener's second attempt at AI safety legislation, following Governor Newsom's veto of a more expansive bill, SB 1047, last year.
