California Governor Gavin Newsom has signed SB 53, a landmark bill establishing the first state-level transparency requirements for large AI developers including OpenAI, Anthropic, Meta, and Google DeepMind. Authored by Senator Scott Wiener, the legislation compels companies to disclose safety protocols, strengthens whistleblower protections, and creates a new system for reporting critical AI safety incidents to the state’s Office of Emergency Services.
The law covers risks such as model-enabled cyberattacks and deceptive autonomous behavior, areas not addressed by the EU AI Act. While Anthropic endorsed the bill, Meta and OpenAI lobbied against it, warning of regulatory fragmentation. Newsom emphasized that California intends to lead in AI oversight while fostering innovation, positioning the state as a national model as other legislatures, including New York’s, consider similar measures.
