California Enacts 18 New AI Laws While Rejecting Controversial Safety Bill
California Governor Gavin Newsom signs 18 sweeping AI bills into law, addressing risks, training data transparency, privacy, and education. The new laws take effect as the state vetoes the heavily debated SB 1047 safety bill.
California Governor Gavin Newsom signs 18 new AI bills into law, establishing some of the most far-reaching generative AI regulations in the United States. This legislative package covers a wide array of pressing issues, ranging from AI risk assessments and deepfake protections to Hollywood's use of digital clones of deceased performers. The new laws arrive shortly after Newsom vetoes the highly contentious SB 1047, a bill that seeks to prevent catastrophic AI disasters but faces heavy criticism from the tech industry.
The newly signed legislation mandates significant transparency and safety measures for AI developers. One key law requires the California Office of Emergency Services to analyze potential threats that generative AI poses to critical infrastructure and mass casualty events by partnering with frontier model companies like OpenAI and Anthropic. Another major bill forces AI providers to publicly disclose detailed information about their training data, including data sources, the number of data points, and the inclusion of copyrighted material, starting in 2026.
Beyond industry regulations, the new laws extend existing state privacy protections to cover personal information exposed by AI systems like ChatGPT. The package also includes an education component that directs the State Board of Education to integrate "AI literacy" into upcoming math, science, and history curricula. By balancing these practical safeguards with technological advancement, California aims to address the immediate risks of AI while maintaining its position as the global hub for artificial intelligence development.