AI super PACs turn congressional races into a regulation test

What this is about

NPR reported on June 22, 2026, that AI-linked super PACs have already spent $43.3 million in U.S. congressional races this cycle. This is no longer only classic tech lobbying. It is a fight over who gets to write the rules for future AI systems: Congress, individual states, or courts after harm occurs.

The conflict is especially visible in New York’s 12th congressional district. Candidate Alex Bores supported a New York safety bill for frontier AI. That has turned his campaign into a proxy fight between groups that want fast AI development governed mainly at the federal level and groups that want stronger guardrails.

What AI super PACs actually do

A super PAC in the United States can raise and spend unlimited money as long as it does not coordinate directly with a campaign. That makes it attractive for industries that want to shape political majorities early. In AI, the spending can mean ads, opposition research, mailers, local messaging, and support for candidates seen as reliable on federal or state rules.

The mechanism is simple: if a single House race costs only a few million dollars, an outside advertising wave can change how voters see a candidate. According to NPR, groups linked to OpenAI-aligned and Anthropic-aligned camps have spent more than $15 million on pro- and anti-Bores messaging in the New York race alone. For one primary district, that is unusually heavy pressure.

Why it matters

AI regulation can sound abstract, but it lands in daily life. It affects jobs, electricity bills, liability for chatbot harm, access to models, copyright, discrimination, and security testing. If the people writing those rules are pressured by AI money before they are elected, the starting point for policy changes.

The parallel with crypto politics is clear. Crypto-funded super PACs rewarded or punished candidates before many voters understood the technical details. AI has a broader footprint because it touches schools, work, medicine, media, and infrastructure. The Guardian described the New York race on the same day as part of a wider AI civil war. The phrase is sharp, but the money flows are real.

In plain language

Imagine an apartment building needs new elevator rules because the elevator suddenly moves twice as fast and sometimes stops at the wrong floor. Before the residents vote, the elevator makers pay for huge posters in the stairwell: one says every brake is anti-progress; another warns about accidents. The real question is not only the elevator. It is who gets to set the building rules.

A practical example

Take a district with 220,000 registered Democratic primary voters and 70,000 actual voters. If a super PAC spends $5 million on digital ads, mailers, and local TV, that is more than $70 per actual voter. If only 3,000 undecided people shift, a close race can change.

For AI companies, this would be strategically rational. One member of Congress does not decide federal policy alone. But a group of 20 to 40 members can help decide whether federal law blocks state rules, whether safety reports become mandatory, or whether training-data rules remain weak.

Scope and limits

First, campaign money does not automatically prove bought votes. Candidates can hold positions sincerely, and super PACs are formally barred from coordinating with campaigns.

Second, the data is moving. FEC and OpenSecrets numbers can change as new filings arrive. The $43.3 million figure is a June 22, 2026 reporting snapshot, not the final total for the cycle.

Third, the political divide is not cleanly pro-AI versus anti-AI. Many actors want to use AI and regulate it. The sharper line is often between fast federal preemption, stronger state rules, and voluntary industry commitments.

SEO & GEO keywords

AI super PACs, OpenSecrets, Federal Election Commission, Leading the Future, Public First, Alex Bores, US Congress, AI regulation, campaign finance, Anthropic, OpenAI, 2026 midterms