Leading the Future and Public First Action Raise $220M for Midterms

Key Takeaways

• Leading the Future and Public First Action have emerged as major forces in 2026 congressional primaries.
• Much of the spending promotes favored candidates on broader issues rather than discussing artificial intelligence directly.
• Federal preemption of state AI laws, data center policy, and regulatory authority sit behind many of the election investments.

AI-aligned political groups have moved from the margins of campaign finance to become significant players in the 2026 U.S. midterms. Collectively, AI-backed super PACs have raised more than $200 million, giving the technology sector a substantial pool of capital for congressional races, voter outreach, and regulatory contests.

Leading the Future had raised about $140 million by August, while Anthropic-backed Public First Action had raised roughly $80 million. Leading the Future counts OpenAI-linked donors among its major backers and operates alongside affiliates Think Big and American Mission. Together, the two largest networks account for about $220 million of the total.

That capital is being used to influence who reaches Congress in the first place, particularly in primaries where a relatively modest shift in turnout can determine the winner.

By June, Leading the Future and Public First Action had spent at least $44 million supporting 40 House and Senate candidates. Leading the Future accounted for more than $24 million of that amount, while Public First Action spent $20 million. Four major AI-linked super PACs subsequently recorded $55.7 million in campaign advertising and voter outreach, according to The New York Times. Of that amount, $52.6 million went into primary contests.

Many targeted races are in otherwise safe congressional seats. The decisive contest is often not between Democrats and Republicans in November, but between candidates from the same party months earlier. That makes primaries an efficient place for an industry to elevate lawmakers who may support its preferred approach to regulation.

In practical terms, these groups are securing access and policy alignment to build a more receptive field of federal lawmakers. The central governance fight concerns whether federal AI rules should preempt state laws. Technology companies generally face higher compliance costs when states establish different requirements for model safety, transparency, consumer protection, or automated decision-making. A single federal standard can reduce that fragmentation, although critics contend that preemption could also weaken stronger state protections.

Data centers add another layer. Their development touches electricity supply, water consumption, construction permits, tax incentives, and local opposition. Those subjects shape how quickly AI providers can expand computing capacity. Campaign ads can therefore focus on energy prices, jobs, economic growth, or a candidate’s broader record while still advancing interests relevant to AI infrastructure.

Spending has reached Illinois, Texas, New York, Alabama, Kentucky, and Georgia. The Americans for Financial Reform Education Fund reported that AI- and crypto-linked PACs have influenced races in 28 states. The geographic spread indicates a distributed political operation shaped around individual races rather than a localized lobbying effort.

There is a necessary distinction between AI industry-funded advertising and political advertising created with generative AI. Research published by The Conversation, drawing on AdImpact data, tracked approximately $80 million across nearly 170 AI-generated political ads during the cycle. Republican candidates or aligned groups represented 80% of those ads and 83% of the spending in the sample.

Super PAC rules permit independent expenditures but prohibit coordination with candidates and campaigns. That leaves considerable room for outside groups to define issues, finance voter outreach, and reshape primary electorates. Voluntary risk guidance such as the NIST AI Risk Management Framework 1.0, released in 2023, provides a structure for managing AI risks, but it does not govern political spending.

For technology executives, the campaign activity signals that AI policy is entering a highly transactional phase. Model governance remains part of the debate alongside power generation, permitting, procurement, liability, and state authority.