California Governor Gavin Newsom (D-Calif.) has vetoed an artificial intelligence (AI) safety bill that would have provided first-in-the-US safety measures for large AI models.
The bill, known as SB 1047, would have mandated third-party testing and whistleblower protections to prevent "severe harm" such as mass casualties or property damage over $500M.
This bill risked stifling innovation, particularly among smaller AI companies and open-source developers. It focused on hypothetical risks rather than practical, targeted safety solutions. Newsom's decision preserves California's leadership in AI by promoting a more balanced, science-based approach to regulation that supports innovation while effectively addressing safety concerns.
Newsom has retreated from a crucial first step toward meaningful regulation of potentially dangerous AI technologies. SB 1047 sought to regulate AI to prevent catastrophic harm while encouraging responsible development. Newsom's veto is a missed opportunity, compromising proactive governance and preserving the status quo.