The AI companies spent $27 million trying to buy influence in a single congressional race and the guy who won told them both to pound sand.
The Summary
- Micah Lasher narrowly won New York's 12th Congressional District Democratic primary after Anthropic and OpenAI effectively cancelled each other out with dueling super PAC spending
- Alex Bores, who co-authored New York's RAISE Act requiring AI safety guardrails, lost despite being boosted by backlash to the anti-regulation attack ads targeting him
- Lasher used his victory speech to promise AI regulation anyway, rejecting both companies' attempts at influence
- This marks the most expensive AI industry political intervention to date and the first clear signal that direct corporate warfare in elections might backfire spectacularly
The Signal
Here's what actually happened in New York's 12th. Anthropic and OpenAI, through the pro-AI super PAC "Leading the Future," spent $27 million in a Democratic primary to replace retiring Rep. Jerry Nadler. The target was Alex Bores, a former tech employee turned state assemblyman who had the audacity to pass actual AI regulation. Bores co-authored the RAISE Act, which mandated safety requirements and guardrails for frontier AI companies. New York signed a version into state law last year.
The super PAC's strategy was textbook political intimidation: flood the zone with ads painting Bores as anti-innovation and anti-jobs. But they miscalculated the blowback. Bores's popularity surged after voters realized he was being targeted by AI money. In a Manhattan district, getting attacked by Silicon Valley cash plays differently than it does in, say, a swing district desperate for economic development. Voters smelled astroturf.
"The AI companies spent $27 million and the winner promised to regulate them anyway."
Enter Micah Lasher, who wasn't the super PAC's preferred candidate but won the primary by a narrow margin. Instead of playing nice with his accidental benefactors, Lasher used his victory speech to reject both Anthropic and OpenAI and promise regulation. This is the political equivalent of taking someone's money, buying a megaphone, and using it to tell them to get lost.
The math here is brutal for the AI companies:
- $27 million spent in a single House primary
- Zero influence gained with the winner
- Activated voter skepticism about Big Tech money in politics
- Made AI regulation more popular, not less
What makes this particularly fascinating is the proxy war angle. Leading the Future is a $100 million super PAC, meaning this Manhattan primary was just the opening salvo. The fund exists specifically to defend AI companies from regulatory threats by electing friendly candidates and punishing hostile ones. But if your first major intervention produces a candidate who actively campaigns on ignoring you, the deterrent effect collapses.
The Implication
Watch what happens next in races where Leading the Future deploys. If other candidates start using AI super PAC attacks as a campaign asset rather than a liability, the entire $100 million war chest becomes a credibility problem. Smart politicians will now bait the attacks, knowing the backlash works in their favor with educated urban voters.
For AI companies, this is a master class in why you don't do politics like a product launch. You can't A/B test your way into Congress. You can't buy influence the way you buy user acquisition. Lasher's victory speech was a signal: the rules of engagement just changed, and the old Valley playbook of spending your way to friendly regulation is dead in districts that actually understand what you're doing.