California’s New AI Regulation: The Frontier AI Act

On Monday, California Governor Gavin Newsom signed a significant bill that mandates artificial intelligence companies to share safety information regarding their advanced frontier models. This move is seen as a substantial step toward regulating AI in Silicon Valley.

The Frontier AI Act, known as SB 53, requires developers of leading AI models to create and publish a framework that outlines how to evaluate and mitigate severe risks associated with their technologies.

Governor Newsom made a statement expressing optimism about balancing the growth of the AI sector with the need for community protection. He emphasized that California is setting a precedent with a focus on public trust as this rapidly evolving technology develops.

Recently, Newsom indicated his support for the bill, marking a key shift after previously rejecting earlier versions of AI legislation.

SB 53 builds on last year’s SB 1047, which aimed to mandate safety testing of AI models before their release and hold developers accountable for any resulting harm. Though SB 1047 successfully passed the state legislature, it was ultimately vetoed by Newsom.

While some California lawmakers face skepticism from Congress regarding various laws, SB 53 has garnered broader support.

Jack Clark, co-founder and policy director of an AI company, voiced his endorsement of the bill, stating it establishes essential transparency for frontier AI companies without imposing rigid technical requirements.

Clark noted that, despite the importance of federal standards to prevent a confusing patchwork of state regulations, California’s framework effectively balances public safety with technological innovation.

Experts believe that California’s influential position in the AI industry could help pave the way for national standards, especially as federal progress remains sluggish.

However, potential opposition looms from the Trump administration and some Republican lawmakers who have expressed concerns over California taking the lead in establishing AI regulations that may stifle innovation nationwide.