California Governor Gavin Newsom has vetoed Senate Bill (SB) 1047, citing concerns that the bill’s regulatory scope was overly broad and lacked sufficient empirical support. Newsom’s decision, announced on Sunday, marks a significant development in the state’s approach to artificial intelligence (AI) legislation. SB 1047 aimed to impose stringent standards on large AI models, but Newsom argued that these regulations, while targeting expensive and large-scale systems, might give the public a false sense of security. He emphasized that smaller, specialized models could pose equally significant risks as their larger counterparts, thus necessitating a more nuanced regulatory framework.
The bill, introduced by Senator Scott Wiener, had successfully passed the California State Assembly in late August with a 49-15 vote. This legislative initiative received backing from notable figures such as Tesla CEO Elon Musk. At the same time, it faced criticism from policymakers, including Congresswoman Nancy Pelosi, who voiced concerns about the bill’s potential negative impact on innovation and the U.S. AI ecosystem (JURIST).
Despite the veto of SB 1047, Governor Newsom has recently signed three other AI-related bills into law, focusing on mitigating election-related content generated by AI, particularly deepfakes. These legislative actions underscore California’s leadership in AI regulation and its responsibility, being home to 32 of the world’s top 50 AI companies. Newsom has also announced future initiatives for safeguarding Californians against AI misuse, highlighting ongoing collaborations with experts like Dr. Fei-Fei Li (JURIST).
The divergence in views on AI regulation continues to shape the debate globally. SB 1047’s framework, which focused on the cost and scale of AI models, contrasts with perspectives that advocate for regulations grounded in the actual risks posed by AI systems, independent of their size and computational expense.