Nancy Pelosi’s replacement is ready to regulate AI

Scott Wiener, a California state senator, has emerged as a frontrunner to regulate artificial intelligence (AI) in the wake of former Speaker of the House Nancy Pelosi's congressional victory in San Francisco. Wiener, who authored legislation aimed at curbing AI risks during his time as a state senator, now stands poised to play a pivotal role in shaping federal AI policy.

In the wake of a report from AI nonprofit METR that suggested Anthropic had achieved recursive self-improvement, Wiener expressed deep concern about the rapid pace of AI development and the potential for a catastrophic AI-related disaster. He emphasized the need for decisive action by Congress to address these concerns before it's too late.

Wiener's focus on AI regulation has been growing in recent years, and he has already made a significant impact on AI policy in California. However, his efforts were thwarted when Governor Gavin Newsom vetoed a bill he authored. Despite this setback, Wiener has continued to advocate for AI regulation, drawing on his experience as a state legislator to push for stronger federal oversight.

During his congressional campaign, Wiener received support from Public First Action, a nonprofit focused on AI regulation that has received substantial funding from Anthropic. While some have criticized him for being too close to the AI industry, Wiener maintains that his advocacy is driven by his long-standing commitment to AI safety, not financial ties to the industry.

If Congress establishes a select committee on artificial intelligence, Wiener would be eager to serve. He supports policies such as strong federal regulation, whistleblower protections, testing and auditing requirements, and a kill switch mechanism to address AI risks. Wiener also opposes broad federal preemption of state AI laws, emphasizing the importance of states maintaining their flexibility to experiment with AI regulation.

Wiener's optimism about Congress's ability to rein in AI is cautiously tempered by the recent failure of several anti-data-center ballot measures, which he believes reflects a broader public concern about the local impacts of AI rather than the broader risks associated with AI development. Despite this, Wiener remains hopeful that Congress will take meaningful action to address AI risks in a timely manner.