OpenAI Backs Proposed California Rules for AI Model Shutdowns and Audits

OpenAI is urging California to strengthen SB53 with proposed requirements for AI companies to add model shutdown mechanisms, undergo annual independent audits and bolster cybersecurity protections.

Chan Park, OpenAI’s head of U.S. and Canada policy and partnerships, signed an Oct. 20 letter to Gov. Gavin Newsom recommending stronger rules under the law. The letter also supports requiring companies’ safety frameworks to address frontier models’ ability to conduct AI research and development autonomously.

California passed SB53 in 2025. The law requires frontier AI companies to publish safety frameworks and transparency reports and report major safety incidents.