The Pentagon has threatened to cancel Anthropic’s contract by Friday if the company does not agree to the department’s terms for the use of its AI model, sources confirmed to The Hill on Tuesday.
Defense Secretary Pete Hegseth met with Anthropic CEO Dario Amodei on Tuesday at the Pentagon amid a dispute over the AI firm’s usage policy, which bars its model Claude from being used for mass surveillance or to develop weapons that can be used without human oversight.
If Anthropic doesn’t agree to the Pentagon’s terms, the department warned it would use the Defense Production Act against the company or designate it as a supply chain risk, a senior Pentagon official told The Hill. Axios first reported the Friday deadline.
“During the conversation, Dario expressed appreciation for the Department’s work and thanked the Secretary for his service,” an Anthropic spokesperson told The Hill in a statement on Tuesday.
“We continued good-faith conversations about our usage policy to ensure Anthropic can continue to support the government’s national security mission in line with what our models can reliably and responsibly do,” the spokesperson added.
The Pentagon official insisted that tactical operation cannot be led by exception and the legality of the missions are the department’s responsibility as the end user.
“The Pentagon has only given out lawful orders,” the official said.
The relationship between Anthropic and the Pentagon has become increasingly rocky in recent weeks, putting at risk the $200 million contract the company signed with the Defense Department last summer.
Google, OpenAI and xAI also struck similar contracts and have since been added to the department’s new bespoke AI platform, GenAI.mil.
However, Anthropic’s Claude has so far been the only AI model available on classified systems. The Pentagon reached a new agreement with Elon Musk’s xAI to use its AI model, Grok, on the classified side, the Pentagon official told The Hill, while Google and OpenAI are “close.”
Despite the months-long back-and-forth, Tuesday’s meeting was respectful and cordial and both sides were thoughtful and friendly with no one raising their voice, a source familiar with the meeting told The Hill.
Amodei underscored that no one operating in the field has encountered issues with the company’s two red lines, mass surveillance and lethal autonomous weapons, according to the source.
However, the Pentagon official said that “this has nothing to do with mass surveillance and autonomous weapons being used.”
Amodei also noted that Anthropic had no outreach to the department or Palantir after the Jan. 3 raid that resulted in the capture of Venezuelan President Nicolas Maduro and that the AI startup has never objected or interfered with “legitimate” military operation, the source familiar with the meeting said.
Previously, a senior Pentagon official told The Hill that a senior Anthropic executive was talking with a senior Palantir executive and asking if Anthropic’s AI model was used during the early January operation.
The Palantir executive told the Pentagon about the interaction since he was alarmed the question was brought up in way that would signal that Anthropic could disapprove of use of its model.
The Defense Production Act, which was enacted 1950, gives the president broad authority to control domestic industries for the purpose of national defense. The law was used during the COVID-19 pandemic to boost the production of vaccines.
The Pentagon official said the department would use the Defense Production Act to compel Anthropic to have its model used however the department sees fit, regardless of whether the AI firm wants to or not.
https://thehill.com/policy/defense/5752960-pentagon-threatens-anthropic-contract/So the trump administration agreed to and signed a contract with the current constraints in place. But now, mid contract they expect to change the terms and parameters of the contract or they will cancel it? They have officially adopted the trump business model.