$300 Million to Buy Congress: How Big AI Is Secretly Rewriting America's Political Rules
Silicon Valley Buys the Midterms
The AI industry has committed $300 million to shape Congress. This isn't a tech story. It's a referendum on who writes the rules of the future.
The ads flooding New York's 12th Congressional District don't mention artificial intelligence. They talk about immigration. Healthcare. Safety. But behind the imagery, a super PAC tied to the AI industry has quietly spent more than $2.2 million to defeat a single Manhattan assemblyman — Alex Bores, a former Palantir employee who had the audacity to co-sponsor a state AI accountability bill.
Welcome to the 2026 midterms, where the most consequential political arms race in American tech history is unfolding — and where the real stakes aren't the candidates. They're the rules.
$300 Million and Counting
The AI industry and its affiliated executives have now committed over $300 million to federal and state races this cycle — the single largest tech-sector political intervention in U.S. history. The money flows across party lines through rival PACs with opposing visions, targeting lawmakers not by party but by a single variable: their stance on AI regulation.
The dominant force is Leading the Future, backed by OpenAI co-founder Greg Brockman and his wife Anna, who each contributed $12.5 million, along with venture capitalist Marc Andreessen and Palantir co-founder Joe Lonsdale. The group raised $125 million in the second half of 2025 alone and entered 2026 with $70 million cash on hand, pledging to spend at least $100 million to elect lawmakers who would establish a single federal AI framework — and crucially, override state-level restrictions.
In late March, a second pro-deregulation operation entered the field. Innovation Council Action, championed by Trump White House AI advisor David Sacks, announced plans to spend at least $100 million more, backed by a congressional scorecard that ranks lawmakers by alignment with the administration's agenda. Meta has separately earmarked roughly $65 million, primarily for state contests.
Against them stands Public First Action, a 501(c)(4) founded by former lawmakers and funded with a $20 million pledge from Anthropic — the maker of Claude and one of the few AI companies publicly arguing that the technology poses "considerable risks." The group has raised approximately $50 million total, running ads praising legislators like Sen. Marsha Blackburn and Rep. Josh Gottheimer for supporting safety measures.
The industry's internal divide is real. But from Wall Street's perspective, it may not matter as much as it seems.
A Price Discovery Event, Not a Policy Debate
The conventional read is that AI regulation risk is rising. The truer read is almost the opposite.
The White House's March 20 National Legislative Policy Framework for AI explicitly calls for federal legislation that would preempt state AI laws. If Congress delivers, companies gain what they want most: one permissive national sheriff instead of fifty unpredictable ones. Markets historically prefer that outcome. Scaled incumbents — frontier labs, hyperscalers, semiconductor leaders, data-center infrastructure — stand to benefit most. Compliance written once, in Washington, by firms with lobbying muscle and legal teams, becomes a moat, not a burden.
The losers are less visible. Startups don't get banned under a "deregulatory" federal framework; they get buried in reporting standards, model evaluations, and procurement hurdles that giants absorb and they cannot. States like California that resist federal preemption face constitutional confrontation. And the "AI is just another software story" investment thesis is quietly expiring.
The Darker Angle
Early primary results suggest the money is working. Of roughly 20 AI-backed candidates in Texas and North Carolina, most won. Illinois showed PAC money isn't omnipotent — but it is influential. In Texas alone, AI-aligned groups have spent more than $2.8 million in congressional races, almost entirely backing Republicans.
Rep. Alexandria Ocasio-Cortez has called the money "toxic." Experts counter that it is democracy functioning in plain sight.
Both are right. This is the moment AI stops being priced as technology and starts being priced as regulated infrastructure — joining railroads, oil, telecom, and banking as industries where returns are shaped as much by political architecture as by product quality.
The market, for now, should probably rally on that.
Democracy probably shouldn't.
not investment advice
Sources: Anthropic's $20M to Public First Action — The Hill & LA Times: https://thehill.com/policy/technology/5735472-anthropic-donates-20m-public-first/
Greg Brockman / Leading the Future — Wikipedia & Quartz: https://en.wikipedia.org/wiki/Leading_the_Future