California Governor Vetoes AI Safety Bill as Industry Stresses Innovation

California Governor Vetoes AI Safety Bill

California Governor Gavin Newsom recently vetoed the proposed Safe and Secure Innovation for Frontier Artificial Intelligence Models Act (SB 1047). This bill was set to introduce the most stringent AI legal framework in the United States, focusing on a variety of safety protocols for AI companies.

Key Provisions and Backlash

Initially presented in late August, SB 1047 aimed to impose rigorous requirements on AI models costing over $100 million to train. It included mandates for developers to implement critical safeguards, such as a “kill switch” to mitigate potential disasters including cyberattacks or pandemics.

• The bill included protections for whistleblowers reporting violations.
• Amendments stripped the proposal for a new regulatory agency.
• Companies objected to the bill, arguing it could stifle innovation.

Mixed Reactions from Tech Leaders

Influential voices such as OpenAI's Jason Kwon and Anthropic's Dario Amodei commented on the legislation's implications. Kwon cautioned that it may slow AI progress, while Amodei noted improvements following amendments.

Political Backing and Opposition

Prominent figures like former House Speaker Nancy Pelosi and San Francisco Mayor London Breed opposed the bill. Conversely, supporters included Elon Musk and various entertainers, highlighting a divided public sentiment.

Future of AI Regulation

The federal government is also exploring AI regulation, with recent Senate proposals addressing the impact of AI on critical sectors.