Anthropic forms political team to win support for 2028 candidates, amid dispute with Trump

Anthropic formed an internal political team to provide AI policy advice to 2028 presidential candidates

Drawing lessons from the turbulent relationship with the Trump administration, Anthropic decided to start building connections with the next presidential candidate. The move was seen by critics as an attempt by the company to interfere in politics. At the same time, the company is still fighting a legal battle with the Trump administration in court.

Will Anthropic provide advice to the Trump administration?

Anthropic recently posted a job search for a "Political Project Supervisor" position on its recruitment website, based in Washington, DC, with an annual salary ranging from $295,000 to $345,000. The job description states that the position will be responsible for "building and operating Anthropic's presidential interactive program for the 2028 election cycle," combing through potential candidates from both parties, and providing briefings on artificial intelligence policies to the campaign and its advisers.

Katie Harbath was Facebook's public policy director for ten years and currently runs the consulting firm Anchor Change. She said the move should be seen as the company's attempt to build a "relationship" with the person who will enter the White House in 2029. She believes the role itself is not uncommon and compares it to her past work, although she notes that it is rare to include a company's political action committee funds in the same combination.

Anthropic's motivation for establishing this relationship may be related to its confrontational dealings with the Trump White House.

In March, the U.S. Department of Defense listed Anthropic as a "supply chain risk" and ordered federal agencies to stop using its Claude model because Anthropic refused to let the Pentagon use its technology to develop fully autonomous weapons or conduct large-scale surveillance of Americans. The company subsequently filed a lawsuit and filed two complaints with the Ministry of Defense. In the California case, U.S. District Judge Rita Lin ruled in August that the designation violated the First Amendment, was unlawful retaliation, and was "arbitrary and capricious."

However, on September 25, the U.S. Court of Appeals for the District of Colombia Circuit upheld the appointment in a 2 - 1 decision. Defense Secretary Pete Hegseth posted on Platform X: "Confirm: @AnthropicAI = Supply Chain Risk."

CEO Dario Amodi has always advocated slowing the development of artificial intelligence and strengthening government regulation, a position opposed by President Trump. On September 29, Amodi signed a voluntary self-regulation agreement with other artificial intelligence company executives at the White House.

How does Anthropic fund its political projects?

Anthropic registered its employee-funded political action committee, AnthroPAC, with the Federal Election Commission in April, where employees can donate up to $5,000 a year, which is the legal limit.

Federal Election Commission documents show that from April registration to the end of August, the political action committee raised $372,308, spent $156,749, and had approximately $215,559 in cash holdings.

Anthropic spends much more on non-political action committee channels. It donated $20 million to Public First Action, a bipartisan group linked to the super political action committee of two candidates who support artificial intelligence security measures. In addition, Amodi personally donated US$1 million to the relevant pro-regulatory super political action committee Public First in July this year.

Some observers, such as Bruce Freed of the non-partisan Center for Political Accountability, view the project as a lobbying attempt, calling it "an artificial intelligence company effort to shape or prevent government regulation or oversight."

Senator Alex Padilla, Democrat of California, urged the company to disclose more information, warning that "education and influence are different, and the public has a right to know where the line lies."