Scott Wiener, Author of California AI Laws, Wins Pelosi’s House Seat
The California senator behind SB 1047 and SB 53 heads to Congress, saying he wants a strong federal AI regulator and opposes broad preemption of state laws.
Scott Wiener, the California state senator best known nationally for writing some of the country’s most closely watched artificial intelligence safety bills, has won the race to succeed former House Speaker Nancy Pelosi in San Francisco. His victory sends one of the most prominent state-level AI regulators to Washington just as Congress debates whether, and how, to set national rules for the technology.
What happened
Wiener defeated Connie Chan, a member of the San Francisco Board of Supervisors, in California’s 11th Congressional District, which covers all of San Francisco. Chan conceded on Wednesday, a day after the polls closed. According to the San Francisco Standard, the latest Department of Elections tally released that afternoon showed Wiener with 55.5% of the vote to Chan’s 44.5%, with about 164,000 ballots still left to count. KTVU, citing The Associated Press, reported similar figures, roughly 55% to 44%, with about 57% of the vote counted.
In her statement, Chan said her campaign expected the margin to narrow but had concluded that “we do not have a clear path to victory,” and she congratulated Wiener. Wiener thanked Chan for what he called a “spirited and hard-fought campaign.”
The race was notable for several reasons. Pelosi, who has represented San Francisco for nearly four decades, endorsed and raised money for Chan. After the result, the Standard reported, Pelosi praised Chan’s campaign, congratulated Wiener and offered to help with the transition. Wiener, who is openly gay, will be the first LGBTQ person to represent San Francisco in Congress when he takes office in January.
It was also expensive. The Standard reported that Wiener and groups backing him spent about $15.5 million, compared with roughly $4.6 million by groups supporting Chan. MIT Technology Review reported that Wiener’s campaign received more than $3 million in support from Public First Action, a nonprofit that backs AI regulation and gets substantial funding from the AI company Anthropic. Opponents argued that the support showed Wiener was too close to the AI industry. Wiener rejected that view, telling the publication the group backed him because he was the first lawmaker to introduce and pass AI regulation that actually took effect.
Why it matters for AI policy
Wiener’s 2024 bill, SB 1047, would have imposed safety obligations on developers of the most powerful “frontier” AI models. It drew fierce debate in Silicon Valley and was vetoed by California Governor Gavin Newsom. Its lighter-touch successor, SB 53, which focuses more on transparency from large AI developers, was signed into law. MIT Technology Review noted that both bills have inspired proposed federal legislation.
In an interview with MIT Technology Review the Friday after the election, Wiener laid out what he hopes to push for in Congress:
- A “strong, empowered, nimble federal regulator” for AI, pointing to a proposal from Rep. Ro Khanna, a California Democrat.
- Protections for whistleblowers at AI companies.
- Testing and auditing requirements for advanced AI systems, along with a “kill switch” requirement.
- Using congressional oversight to hold AI labs accountable.
He also said he opposes broad federal preemption, meaning a national law that would override state AI rules. He argued that states act as policy laboratories, a reference to the idea associated with former Supreme Court Justice Louis Brandeis. Preemption has been one of the most contested issues in Washington’s AI debate, with some in the tech industry favoring a single national standard over a patchwork of state laws.
Wiener said he is concerned about the risk of humans losing control over advanced AI. He pointed to a report from METR, a nonprofit that evaluates AI models, which was released the day before the interview and was widely read as evidence that Anthropic had achieved what researchers call recursive self-improvement, in which AI systems help improve themselves. “It’s very concerning,” he said, adding, “We need to slow this down.”
The political landscape
Wiener told MIT Technology Review that he had hoped Congress would take decisive action on AI before the end of this year, but that this no longer looks likely. If House Democrats create a select committee on AI, an idea the publication said they have discussed since September, he said he would be eager to serve on it. He described himself as “cautiously optimistic” about Congress’s ability to regulate the technology, given its mixed record on tech rules.
He also acknowledged obstacles. President Donald Trump called concerns about catastrophic AI risk a “hoax” in September, according to MIT Technology Review, and Wiener said getting legislation past the White House would be a challenge. He added that cooperation with China on AI safety is essential.
Wiener’s views on AI’s economic effects have also drawn attention. KTVU reported that he has said corporations should be taxed on profits they gain by replacing human workers with AI. The campaign itself showed how fraught the technology can be in politics: in August, Wiener launched an AI chatbot parody of Chan, promoted with billboard ads, that Pelosi and other Democratic women condemned as a “gross misuse of technology,” according to KTVU. Critics also called it racist and sexist. Wiener apologized and took the website down.
What comes next
Wiener will be sworn in in January as part of a new Congress in which Democrats will have more power to conduct oversight. Public concern about AI’s local effects is also growing. MIT Technology Review noted that voters approved 40 of 41 ballot measures opposing data centers on Election Day. Wiener said he sees that backlash as driven mostly by local concerns such as electricity bills, water use and neighborhood impacts, rather than fears about AI risk.
