california governor newsom signs landmark ai safety bill sb 53

California Enacts Landmark AI Transparency Law

Governor Gavin Newsom of California has approved SB 53, a groundbreaking piece of legislation establishing unprecedented transparency standards for prominent AI companies operating within the state.

New Requirements for AI Labs

SB 53, recently passed by the state legislature, mandates that substantial AI laboratories – encompassing entities like OpenAI, Anthropic, Meta, and Google DeepMind – disclose their safety protocols. Furthermore, the bill provides crucial protections for employees who act as whistleblowers within these organizations.

A key component of SB 53 is the creation of a reporting system. This allows both AI companies and the general public to submit reports concerning potentially critical safety incidents to California’s Office of Emergency Services.

Reporting Obligations and Scope

AI companies are also obligated to report incidents involving crimes perpetrated without human intervention, such as cyberattacks. Additionally, they must document instances of deceptive model behavior that fall outside the scope of regulations like the EU AI Act.

Industry Response and Lobbying Efforts

The reception to this bill within the AI industry has been varied. While some firms have expressed concerns about a fragmented regulatory landscape hindering innovation, Anthropic publicly supported the legislation.

Notably, Meta and OpenAI actively lobbied against SB 53. OpenAI even published an open letter addressed to Governor Newsom, urging him to veto the bill.

Financial Influence and Political Action

Recent months have seen significant financial contributions from Silicon Valley leaders to super PACs supporting candidates who favor a less restrictive approach to AI regulation. Leaders from OpenAI and Meta have spearheaded the formation of pro-AI super PACs, aiming to support favorable candidates and legislation.

Potential for National Impact

Other states may now consider California’s actions as a model as they seek to mitigate the potential risks associated with the rapid development of powerful AI technologies. A similar bill has already been passed in New York and awaits Governor Kathy Hochul’s decision.

Governor Newsom's Statement

“California has demonstrated its ability to implement regulations that safeguard our communities while fostering the continued growth of the AI industry,” Newsom stated. “This legislation achieves that balance. AI represents a new era of innovation, and California is positioned as a national leader by enacting the first-in-the-nation frontier AI safety legislation, building public trust as this technology evolves rapidly.”

Further Legislation Under Consideration

Governor Newsom is also currently reviewing SB 243, a bill that received bipartisan support in both the State Assembly and Senate. This bill focuses on regulating AI companion chatbots, mandating safety protocols and establishing legal accountability for operators if their bots fail to adhere to these standards.

Legislative History and Compromise

SB 53 represents Senator Scott Wiener’s second attempt at AI safety legislation. His previous, more extensive bill, SB 1047, was vetoed by Newsom last year following substantial opposition from AI companies. For this bill, Wiener proactively engaged with major AI companies to address their concerns and incorporate feedback.