The AI Industry’s First Midterm: Who’s Spending, and What They Want

by

For the first time, the artificial intelligence industry is spending on a midterm election at the scale of the biggest players in Washington. AI-linked political groups had raised more than $107 million for federal races by mid-August (Newsweek). Unusually, the industry is not spending as a single bloc. Its money is split between rival camps that disagree about how tightly AI should be regulated, and in some races they have spent millions against each other.

The three camps

Camp Main backers Money What it wants
Leading the Future network Andreessen Horowitz and its founders, OpenAI president Greg Brockman, Joe Lonsdale, Ron Conway, Perplexity AI More than $125 million raised (Newsweek) Rapid AI development and lighter regulation (Axios)
Public First network Anthropic, plus $1 million from its CEO, Dario Amodei About $80 million raised, including $40 million from Anthropic (Newsweek) AI safety, transparency and stronger oversight (Axios)
Meta-backed groups Meta $45 million to one national group and $65 million pledged to two others (Axios) Fighting state AI bills it considers burdensome (Axios)

This site uses AI tools, including Anthropic’s Claude, in research and drafting.

Each network mirrors the Fairshake model the crypto industry built: several linked committees aimed at different audiences. Leading the Future works through Think Big PAC, American Mission PAC and a nonprofit, Build American AI, which does not disclose its donors (Axios). Think Big alone had spent $22.4 million by October 5 (OpenSecrets). Public First's political arms include Jobs and Democracy PAC, which backs Democrats, and Defending Our Values PAC, which backs Republicans; the network is led by two former members of Congress, Democrat Brad Carson and Republican Chris Stewart (Axios).

The race that defined the fight

The two camps collided most directly in New York's 12th District, the Manhattan seat left open by Rep. Jerry Nadler's retirement. One candidate, Assemblymember Alex Bores, had sponsored New York's RAISE Act, which requires major AI developers to disclose their safety protocols and report serious misuse (TechCrunch).

Leading the Future spent about $8 million to defeat him. Groups supporting him spent roughly $20 million, most of it from the Anthropic-backed network; Jobs and Democracy PAC alone spent $15 million (Gizmodo, Newsweek). Outside money in the eight-candidate Democratic primary topped $40 million.

Neither side got the clean win it wanted. Bores lost on June 23 to Assemblymember Micah Lasher, who had drawn less outside attention. "Though we've come up short tonight, the example set here was not the one AI oligarchs intended," Bores said. Lasher's message to both camps: "I won't be taking my cues from either of you" (Gizmodo).

Spending in safe seats, with ads that skip AI

The industry's money has gone overwhelmingly into primaries for seats that are not competitive in November. According to a New York Times analysis, OpenAI- and Anthropic-aligned groups had spent $55.7 million by late September, and $52.6 million of it went to safe seats. None of the 95 ads the two networks placed mentioned data centers, and only some mentioned AI at all (AI Weekly, summarizing the Times). The Brennan Center notes that the two companies' groups are backing candidates aligned with each company's agenda in races unlikely to change party hands (Brennan Center).

The logic is the same one the crypto industry uses: the decisive moment in a safe seat is the primary, and a member of Congress who wins one may serve for years. Critics argue that this lets industries choose lawmakers without voters hearing what the industry wants from them.

Signs of strain

In late September, the New York Times reported that Greg Brockman was pulling back from Leading the Future. He and his wife had pledged $50 million and given $25 million, but would not give the rest. Reasons reported include public backlash against the group's attacks on candidates who favor tighter AI rules and limits on data centers, an open letter from Democratic campaign staff urging the party to refuse the group's money, and friction inside OpenAI. OpenAI has said Brockman's involvement was "in a personal capacity, not on behalf of the company" (Gizmodo).

Both sides' case

Leading the Future and its backers say they want rapid AI development and lighter regulation, and Meta-backed groups have focused on fighting state AI bills they consider burdensome (Axios). Public First and its backers argue that powerful AI systems need basic safety and transparency rules now, and that the industry's biggest companies should not be the ones writing them. Both camps say they are bipartisan, and both have backed candidates in each party.

What the two have in common is the method: large sums from a few companies and executives, steered into races most voters will never see as competitive. With the AI industry's spending now rivaling crypto's, and voters telling pollsters they are worried about corruption and the influence of wealthy donors (Brennan Center), how candidates handle AI money may become a campaign issue in its own right, as crypto money already has.

For more on how industry money is shaping this election, see Crypto's $200 Million Bet and Where Has MAGA Inc.'s Money Gone?, whose donors include AI executives.

Sources