Anthropic at loggerheads with US military over AI use

AI giant Anthropic is at risk of losing a lucrative Pentagon contract over its policy that its technology should never be used for the mass surveillance of US citizens or in fully autonomous weapons systems.

A US official told AFP on Monday that Anthropic executives would meet with Pentagon officials on Tuesday to discuss the matter, with a $200 million contract on the line. The meeting was first reported by Axios.

Anthropic is one of the world's most advanced AI companies, alongside OpenAI, the maker of ChatGPT, and Google, whose powerful Gemini models rival both.


The three companies, along with Elon Musk's xAI, were contracted last year to supply AI models for a wide variety of military uses -- including what the Pentagon called frontier AI projects.

As those projects have been deployed, the Pentagon has insisted that all four companies make their models available for any usage that falls within US law.

Anthropic has pushed back, drawing red lines around certain specific applications.

That stance has drawn an angry response from military officials, with reports that the Pentagon could go as far as banning Anthropic altogether from future defense contracts.

"You can't have an AI company sell AI to the Department of War and [then] don't let it do Department of War things," Emil Michael, a senior adviser who functions as the US military's de facto chief technology officer, told specialist outlet DefenseScoop.

The Trump administration began referring to the Department of Defense as the Department of War following an executive order by the president. Formally changing the department's name, however, would require an act of Congress.

Anthropic was founded by former OpenAI employees in 2021 on the premise that AI development should prioritize safety - a philosophy it has maintained even as it looks to deploy its technology across all sectors of the economy, competing directly with OpenAI and Google.

In a statement, Anthropic declined to comment on its reported red lines, but said it was "committed to using frontier AI in support of US national security," adding that it had held "productive conversations, in good faith, with DoW on how to continue that work and get these complex issues right."

According to Axios, defense officials say negotiations are on the verge of breaking down.

The dispute is the latest friction between Anthropic and the Trump administration, whose officials have accused the company of advancing a doomer attitude toward AI and unnecessarily stoking public fears about the technology.