Key points to remember:
- Anthropic filed FEC documents on April 3, 2026, launching AnthroPAC with employee contributions capped at $5,000 per year.
- AnthroPAC follows Anthropic’s $20 million donation to Public First Action in February, moving from championing a cause to funding candidates.
- The AI industry contributed approximately $185 million to the 2026 midterm races, with Anthropic now playing a direct role through AnthroPAC.
Anthropic forms first PAC as AI industry invests $185 million in 2026 midterm races
The committee has FEC ID C00946111 and is classified as a separate segregated fund connected to Anthropic PBC, headquartered at 548 Market Street in San Francisco. Allison Rossi is Treasurer and Keeper of Records. Jared Powell serves as Deputy Treasurer. JPMorgan Chase is listed as the committee’s bank.
AnthroPAC is funded exclusively by Anthropic employees. Federal law caps individual contributions at $5,000 per person per year. The company itself does not contribute directly. All donations and expenditures will be disclosed in FEC records.
A bipartisan board of directors oversees the PAC. Its stated goal is to support current Washington DC lawmakers and emerging candidates from both parties who are active in artificial intelligence policy. The committee’s contact address is PAC@anthropic.com.
The filing comes two months after Anthropic committed $20 million to Public First Action, a bipartisan 501(c)(4) organization working on AI education and federal governance, according to The Hill’s report. At the time, Anthropic said it wanted to support candidates who understand the challenges of reshaping labor markets, national security and global competition through artificial intelligence (AI).
AnthroPAC is shifting this strategy from issue advocacy to direct support of candidates. The distinction is important under federal law. Employee contributions to the PAC go toward individual campaigns, while outside groups like Public First Action fund ad delivery and broader voter outreach.
Anthropic has been specific about the policies it wants to promote. The company has publicly supported model transparency requirements, federal AI governance frameworks that do not fully preempt state laws, targeted export controls on AI chips, and rules focused on high-risk applications.
These positions have created friction with the current administration. Anthropic prevents Claude from being used in lethal fully autonomous weapons or mass surveillance of Americans. The Pentagon responded by calling Anthropic a supply chain risk and suspending or canceling contracts, including a $200 million opportunity. Anthropic sued the Department of Defense. A federal judge has since issued a temporary block on the punitive measures.
The AI industry as a whole has increased its political spending ahead of the 2026 midterm elections. Companies including Google, Microsoft, Amazon and Meta have collectively invested around $185 million in the midterm races so far. AnthroPAC fits the standard employee-funded PAC model common in the tech sector, keeping corporate cash funds out of direct campaign contributions while building political reach.
Social media comments following the April 3 filing presented the move as another sign of the AI industry’s growing involvement in U.S. elections. Some voices aligned with the Trump administration have questioned whether a PAC formed by a company in a legal dispute with the Pentagon could credibly claim bipartisanship.
Anthropic has not released a statement specifically on AnthroPAC. The company’s February remarks on the Public First donation remain the clearest public explanation of its policy goals.
Further details on board composition and initial contributions are expected to appear in future FEC reports. As AI regulation moves closer to the center of the 2026 election cycle, the PAC gives Anthropic employees a formal structure to invest directly in the candidates they want to shape federal policy.
