- Anthropic formally registered AnthroPAC with the FEC on April 3, 2026, establishing a corporate political action committee
- The move comes after the company already funneled $20M into a Super PAC and is suing the Pentagon over its supply-chain risk designation
- The AI lab’s CEO drew Trump’s ire by refusing to let the government use Claude for mass surveillance or autonomous weapons targeting
Anthropic has a message for Washington: if you can’t beat them, fund the people who might. The AI company filed paperwork with the Federal Election Commission on April 3 to officially establish AnthroPAC, a corporate political action committee — a calculated escalation of its political presence in the capital that comes in the middle of an all-out standoff with the Trump administration.
The FEC filing establishes AnthroPAC as a “separate segregated fund” connected to Anthropic PBC, a corporation. Allison Rossi is listed as treasurer. The PAC’s stated aim: bankroll midterm candidates who align with Anthropic’s policy agenda, which now includes surviving an unprecedented government crackdown. The committee can accept contributions up to $5,000 per donor and will support candidates across party lines — a bipartisan charm offensive designed to buy goodwill in a city that hasn’t been especially warm to AI lately.
From Leaked Memos to Federal Bans
The PAC filing is the culmination of months of open hostility between Anthropic and the administration. It started when CEO Dario Amodei publicly refused to let the Pentagon use Claude — Anthropic’s flagship AI model — for mass surveillance of Americans or for autonomously selecting targets to kill. When a memo laying out that position leaked, Amodei apologized. Trump was not moved. He called Anthropic “like dogs” and barred federal agencies from using the company’s technology.
The Pentagon followed up by designating Anthropic a “supply chain risk” — a designation historically reserved for foreign companies linked to adversaries like China.
Rather than fold, Anthropic sued. The company filed federal litigation challenging the supply-chain designation and the government-wide ban, escalating a dispute that has turned the lab into one of the most visible antagonists of the Trump era’s tech nationalism.
Buying Influence While Fighting for Survival
The PAC isn’t Anthropic’s first foray into electoral politics. In February, the company quietly poured $20 million into Public First Action, a Super PAC that has run advertising campaigns supporting AI regulation and candidates friendly to the company’s safety-first worldview. That’s not small change — it’s the kind of money that buys real attention in a midterm cycle.
Industry-wide, AI companies have now committed a staggering $185 million to midterm races, according to the Washington Post. Anthropic’s lobbying spend surged 511% in the first months of Trump’s second term, reaching $3.13 million in 2025. The message from the PAC is consistent with that spending: the company intends to be a permanent fixture in D.C., whether through lawsuits, lobbying, or checkbooks.
The timing is pointed. AnthroPAC’s registration came just days after Trump escalated the government’s campaign against the lab — a signal that Anthropic sees political donations as a necessary complement to legal warfare, not a replacement for it. The administration has argued that unelected tech executives shouldn’t set the terms for how their AI is used by the military. Anthropic’s answer, apparently, is to go around the executive branch entirely and court the people who write the laws.