Introduction

Scott Wiener is headed to the U.S. House of Representatives after winning a competitive election to succeed former Speaker Nancy Pelosi in San Francisco. The California state senator has spent years pushing for artificial intelligence safeguards, and he says federal lawmakers are running out of time to act. In his first comments after the race was called, Wiener pointed to recent research on advanced AI systems as proof that Congress cannot afford to delay.

What Happened

Wiener secured the open congressional seat previously held by Pelosi, prevailing in a closely watched race. Shortly after being declared the winner, he spoke with a journalist about AI risk. He referenced a report from the AI nonprofit METR that some observers linked to Anthropic achieving recursive self-improvement, a scenario where AI systems help improve themselves and potentially accelerate beyond human control. Wiener described the finding as alarming and said he had hoped Congress would pass meaningful AI legislation before the end of the year, though he does not expect that to happen.

As a state lawmaker, Wiener authored SB 1047 in 2024, a bill intended to establish the first major frontier AI rules in the country. California Governor Gavin Newsom vetoed that measure but later signed a narrower bill, SB 53, which Wiener also sponsored. Both pieces of legislation have influenced federal proposals. Wiener's campaign also drew attention for accepting more than $3 million from Public First Action, a group supporting AI regulation with significant backing from Anthropic. Critics argued the funding showed overly close ties to the industry, but Wiener rejected that characterization, saying the group backed him because he introduced and passed AI regulation despite opposition from major technology companies.

Why This Matters

If Congress creates a dedicated AI committee, Wiener said he would want a seat at the table. His policy priorities include a powerful and flexible federal regulator, protections for whistleblowers, mandatory testing and audits of AI systems, and a requirement for a kill switch that could halt dangerous models. He also wants Congress to use its oversight authority to examine AI laboratories.

Voters recently approved 40 of 41 anti-data-center ballot measures, but Wiener interprets that backlash as rooted in local worries about electricity costs, water consumption, and construction near homes rather than a rejection of AI itself. Senator Bernie Sanders has met with AI safety figures and discussed existential risk, potentially giving Wiener an ally in Congress.

At the same time, Wiener opposes broad federal preemption that would block states from passing their own AI laws. Citing the idea of states as laboratories of democracy, he argued that state governments should remain free to experiment with regulations more quickly than Congress can move. He emphasized that any federal action must have real enforcement mechanisms and address the possibility of losing control over advanced AI.

The political environment presents obstacles. President Donald Trump has dismissed catastrophic AI risk as a hoax and has resisted cooperation with China, which Wiener views as necessary to reduce the worst AI dangers. Still, Wiener said he remains cautiously optimistic, partly because Trump's unpredictability could create unexpected opportunities for action—though he hopes it does not take a serious AI incident to force the issue.

Key Takeaways

  • Scott Wiener won the congressional election for Nancy Pelosi's former San Francisco seat.
  • He has made AI safety a central issue, authoring state bills such as SB 1047 and SB 53.
  • Wiener wants a strong federal AI regulator, whistleblower protections, auditing requirements, and a kill switch for dangerous systems.
  • He opposes broad federal preemption of state AI laws and wants states to keep regulatory flexibility.
  • His campaign received over $3 million from Public First Action, an AI regulation group funded in part by Anthropic, prompting criticism he denies.
  • Wiener is cautiously optimistic about Congress but concerned about President Trump's stance on AI risk and U.S.-China cooperation.

Conclusion

Scott Wiener's move from Sacramento to Washington could give the AI safety movement a more prominent voice in federal lawmaking. He enters Congress warning that advanced AI is developing faster than policymakers can respond, and he is prepared to advocate for rules with meaningful enforcement. Whether he can turn those priorities into legislation will depend on a Congress with a mixed record on technology regulation and a president who has publicly downplayed existential AI concerns. For now, Wiener is urging lawmakers to treat the issue with urgency before an avoidable disaster occurs.