California lawmakers to vote on landmark AI regulation bill amidst strong opposition

California legislators are poised to vote on a bill, SB 1047, that aims to regulate the development and deployment of AI across the state. The bill, advanced by State Senator Scott Wiener, would require safety testing for advanced AI models that exceed $100 million in development costs or use significant computing power. Developers would need to implement a ‘kill switch’ for malfunctioning AI and face potential legal action from the state attorney general if they fail to comply. The bill also mandates third-party safety audits and protects whistleblowers.

Despite passing the state Senate with overwhelming support, SB 1047 has faced criticism from key California lawmakers, including Nancy Pelosi and Ro Khanna, who argue that it could drive AI developers out of the state and harm open-source AI projects. They believe the bill’s regulations might create an unfavourable environment for AI innovation in California.

Tech industry leaders have also expressed concerns, with companies like Google, Meta, and OpenAI opposing the bill, fearing it could stifle AI development. They argue that federal regulation would be more appropriate and worry that the bill could negatively impact open-source AI models. However, some notable AI researchers, including Geoffrey Hinton and Yoshua Bengio, have supported the legislation, highlighting the need for robust safeguards in AI development.