Anthropic is spending $20 million to put AI regulation on the ballot. The company announced a donation to Public First Action, a new bipartisan 501(c)(4) organization that plans to back 30 to 50 candidates in state and federal races during the 2026 midterm elections.
The money draws a clear battle line. On one side: Anthropic and Public First Action, pushing for transparency requirements on frontier AI models, federal governance frameworks, and export controls on AI chips. On the other: Leading the Future, a super PAC backed by Andreessen Horowitz, Marc Andreessen, Joe Lonsdale, Ron Conway, and others in the pro-acceleration camp, which plans to spend $125 million stacking Congress with candidates who favor lighter AI regulation.
That is a $20 million vs. $125 million fight, which makes Anthropic the clear underdog in raw spending. But the company is betting that targeted support for candidates who understand AI safety will carry more weight than a broad anti-regulation blitz.
Public First Action, founded by both Republican and Democratic strategists, is organized to work across party lines through affiliated vehicles for each party's primaries. Its policy priorities include giving the public more visibility into how frontier AI companies manage risks, supporting smart export controls to maintain a U.S. advantage over authoritarian competitors, and addressing what it calls the nearest-term high risks: AI-enabled biological weapons and cyberattacks.
Anthropic framed the donation as filling a gap. "At present, there are few organized efforts to help mobilize people and politicians who understand what's at stake in AI development," the company wrote in its announcement. The company also made an unusual concession for a tech firm spending on politics: effective AI governance means more scrutiny of companies like Anthropic, not less.
This is an AI company explicitly lobbying for rules that would constrain its own industry. The cynical read is that regulation favors incumbents - Anthropic can afford compliance costs that smaller competitors cannot. The charitable read is that Anthropic genuinely believes unchecked AI development is dangerous and is putting real money behind that belief. Either way, $20 million makes this more than a press release.