When former Speaker Nancy Pelosi announced her retirement, the fight for California’s 12th district turned into a showdown of policy priorities rather than just party labels. State Senator Scott Wiener and District-1 Supervisor Connie Chan quickly discovered that the most electrifying issue for voters is not housing or transit—it is the looming impact of artificial intelligence. Both candidates claim to want a safer AI future, but their prescriptions differ dramatically, and the contrast is reshaping campaign conversations across the city.
Wiener’s early warning and legislative push
Back in 2021, Wiener began sounding the alarm about the existential hazards of advanced AI systems. By 2024 he had drafted and passed one of the nation’s first state-level AI safety bills a package that required the largest AI firms to install pre-launch safety standards, whistleblower protections and, initially, a “kill-switch” capability. The bill attracted opposition from Governor Gavin Newsom, Rep. Nancy Pelosi and several tech giants, who dismissed the measures as precautionary overreach. Nevertheless, Wiener persisted, trimming the proposal after the veto and securing a watered-down version that Newsom signed in 2025. The final law still obliges companies to adopt emergency shutdown mechanisms and third-party audits—features that have yet to appear in any federal proposal.
Chan’s populist AI pact and tax-focused approach
Chan, a rising progressive figure, signed the national “AI Pact” alongside more than a dozen lawmakers. The pact obliges signatories to oppose tax breaks for data centers, demand safety reviews before major AI products launch, and empower local jurisdictions to set their own rules. During a recent town hall, Chan emphasized that AI must be judged by its effect on working-class families, asking whether the technology will “improve life for working people, or become another tool for billionaires to eliminate jobs and concentrate wealth.” She proposes a moratorium on new data-center construction unless strict conditions are met, and she backs stronger reporting requirements for AI-driven job losses.
Campaign tactics, voter concerns, and the AI showdown
Both campaigns have retooled their canvassing scripts to translate technical jargon into everyday language. Wiener’s field staff now avoid terms like “kill switch” and instead speak of “safety controls,” while Chan’s volunteers stress “protecting workers from automated firings.” After a controversial September incident in which Wiener’s team released a chatbot that mimicked Chan with a negative tone, the candidate publicly apologised, calling the stunt a “mistake.” The episode underscored how quickly AI missteps can become political liabilities. Meanwhile, polls show that San Francisco voters are increasingly asking not just “what is your stance on AI?” but “how will you regulate it?”—a shift that could advantage the candidate perceived as having a concrete, workable plan.
Implications for the congressional race and beyond
If Wiener can persuade enough moderate voters that his pragmatic, bipartisan approach—supporting a new civilian-led federal AI agency while still favoring data-center development with taxes—will keep the technology beneficial, he may preserve the advantage of incumbency that Pelosi once bestowed. Conversely, Chan’s appeal to labor unions and her insistence on tighter federal oversight through existing agencies like the FTC and SEC could resonate with the district’s progressive base, especially if national headlines continue to link AI mishaps to job insecurity. The battle over AI policy is therefore more than a policy debate; it is a litmus test for how San Francisco’s electorate balances tech-industry growth with protecting everyday workers.
As the November election approaches, both campaigns are sharpening their messages, scrutinizing each other’s donation records, and positioning AI safety as a defining issue. Whether the district ultimately favors Wiener’s seasoned legislative record or Chan’s grassroots-focused AI pact could set a precedent for how the United States addresses the rapidly evolving challenges of artificial intelligence at the federal level.
