OpenAI Supports California's Strengthened AI Safety Regulation

Odaily reports: A senior OpenAI official stated in a letter to California Governor Gavin Newsom that the company supports requiring AI companies to implement shutdown mechanisms for their models, undergo mandatory annual independent audits, and adopt stricter cybersecurity measures. The letter, dated Tuesday, was signed by Chan Park, OpenAI's head of policy and partnerships for the United States and Canada. The letter states that OpenAI "recommends strengthening SB53," the existing AI transparency law passed by California in 2025, which currently requires frontier AI companies to publish safety frameworks and transparency reports and to report major safety incidents. Additionally, OpenAI supports regulations requiring companies' "safety frameworks to explicitly cover the capability of frontier models to autonomously conduct AI research and development." (The Information)