Scott Wiener, author of California AI safety bills, wins seat in US Congress

MIT Technology Review reports that Scott Wiener, who wrote California's best-known AI safety bills as a state senator, has won the San Francisco congressional seat being vacated by former House Speaker Nancy Pelosi. In an interview a few hours after the result, he was already focused on advanced AI risk.

Wiener pointed to a report released the previous day by the nonprofit METR, which many read as showing that Anthropic had achieved recursive self-improvement, meaning AI that contributes meaningfully to its own progress. He called it very concerning and said Congress seems unlikely to act decisively before the year ends. His earlier work includes SB 1047, a 2024 frontier AI bill vetoed by Governor Gavin Newsom, and SB 53, a lighter bill that Newsom signed. Both have inspired proposed federal legislation. His campaign received more than $3 million from Public First Action, a pro-regulation nonprofit with substantial Anthropic funding, which opponents used to accuse him of closeness to industry. Wiener rejects that charge.

On policy, Wiener said he would want to serve on a congressional select committee on AI if one is created. He backs an empowered federal regulator, such as the one Representative Ro Khanna has proposed, along with whistleblower protections, testing and auditing rules, a kill-switch requirement and stronger oversight of labs. As a former state legislator he opposes broad federal preemption of state AI laws.

The piece notes that voters approved 40 of 41 anti-data-center ballot measures, but Wiener sees that as driven by local concerns such as electricity bills and water use rather than by worries about AI risk. He calls himself merely cautiously optimistic that Congress can act, and notes that President Donald Trump called catastrophic AI risk a hoax in September.