AnthroPAC Enters the Chat: Anthropic Launches Employee-Funded PAC While Pentagon Drama Heats Up

Anthropic has launched a corporate political action committee, dipping its toes into election financing as AI policy debates rage in Washington. The timing? Chef's kiss. Nothing says "we're just a regular tech company trying to influence policy" quite like filing FEC paperwork while the Pentagon is actively trying to blacklist you.

The company filed paperwork with the Federal Election Commission on Friday to establish "AnthroPAC," an employee-funded PAC that will collect voluntary contributions from staff. Anthropic serves as the "connected organization," with the committee structured as a "separate segregated fund" registered as a lobbyist-affiliated PAC. For those keeping score at home, that's regulatory speak for "we're setting up a perfectly legal way to buy influence while maintaining plausible deniability."

Under US law, individual contributions are capped at $5,000 per election cycle per candidate, with all donations disclosed through public filings. So yes, your friendly neighborhood AI company will be playing in the same sandbox as every other political operator in D.C.—just with more compute and arguably scarier existential implications.

Anthropic says the PAC is expected to support candidates from both major parties. However, some observers have questioned whether the effort will remain politically balanced. Translation: everyone knows "bipartisan" in Washington usually means "we'll fund whoever wins." The PAC equivalent of hedging your bets while pretending to care about democracy.

The timing is interesting. Anthropic is currently facing mounting friction with the Pentagon over the use of its AI systems. In February, the Defense Department designated the firm a supply chain risk after it opposed the use of its technology in fully autonomous weapons and mass surveillance. "Interesting timing" is doing a lot of heavy lifting here—like noticing your house is on fire and deciding now's the perfect moment to start a neighborhood watch.

Anthropic has challenged that designation in court, arguing it reflects retaliation against what it called a protected viewpoint. A federal judge in California has temporarily blocked the measure and paused broader restrictions tied to the dispute. Nothing says "we value your input" quite like the federal government telling an AI company they're a supply chain risk because they had the audacity to say "maybe don't use our tech for killer robots."

The company hasn't been sitting on the sidelines politically this cycle. Anthropic already contributed $20 million to Public First Action, a group focused on advancing AI safety efforts. So to summarize: $20M to AI safety, now a PAC for election influence, and a legal battle with the Pentagon. Anthropic is playing 4D chess while most companies are still figuring out how to not hallucinate in their quarterly reports.

Meanwhile, Google is preparing to support a multibillion-dollar data center project in Texas leased to Anthropic. The project, operated by Nexus Data Centers, could exceed $5 billion in its initial phase, with Google expected to provide construction loans while banks compete to arrange additional financing. Nothing like a $5B data center