California Governor Vetoes Landmark AI Safety Bill;
The Insight Hub, October 1, 2024, 6:o4 pm
California Governor Gavin Newsom has vetoed a state AI safety bill that was considered a significant step towards regulating advanced AI technology in Silicon Valley, with potential global implications.
The bill, passed almost unanimously by the state legislature, sought to mandate safety testing for the most advanced AI systems, known as frontier models, before they could be released. It also empowered the state attorney general to sue companies over serious harm caused by their technologies, such as death or property damage, and required a kill switch for extreme scenarios.
Newsom faced substantial pressure from tech companies and influential members of his own party to reject the bill. While acknowledging the bill’s "well-intentioned" nature, Newsom criticized it for overemphasizing regulation of advanced models without providing adequate means to measure risk or potential harm.
He argued that the bill lacked consideration of whether an AI system was deployed in high-risk environments, involved critical decision-making, or handled sensitive data. Newsom suggested a rewrite involving input from AI experts and business leaders to conduct a science-backed analysis of the risks associated with frontier models.
In the absence of federal legislation, the California bill could have represented one of the first major attempts in the U.S. to regulate AI, similar to efforts underway in the EU and China. The EU's AI Act, which became law this year, relies on mandatory self-regulation of frontier models through codes of conduct, rather than strict formal regulation—a compromise supported by France, Germany, and Italy to protect the bloc's competitiveness.
U.S. tech companies and researchers expressed similar concerns about the California bill. Li Feifei, co-founder of AI startup World Labs, argued in an August opinion piece that the bill would "harm our budding AI ecosystem." Major companies like OpenAI, Google, Meta, and Microsoft also warned that the bill would hinder innovation.
On the other hand, about 50 academics, including AI pioneer Geoffrey Hinton, wrote to Newsom in support of the bill, emphasizing the need for a responsible approach to AI development. "
Comments