Defense Secretary Pete Hegseth is allegedly “close” to cutting off the Pentagon’s ties to AI firm Anthropic and placing it on a blacklist that would force several other firms to stop using the Claude chatbot, according to a report Monday.
Tensions have reached a boiling point after months of heated negotiations, as Anthropic has sought to ensure its tools are not used for mass surveillance on Americans or to develop weapons that can fire with no human involvement, according to Axios.
Hegseth is considering designating Anthropic a “supply chain risk,” meaning any firm that wants to do business with the Department of War will need to cut ties to the fast-growing AI company, a senior Pentagon official told the outlet.
“It will be an enormous pain in the ass to disentangle, and we are going to make sure they pay a price for forcing our hand like this,” the official told Axios.
The “supply chain risk” label is a strict penalty usually reserved for foreign adversaries suspected of posing a national security risk – but senior defense officials have long been frustrated with Anthropic and are eager to pick a fight, a source told Axios.
The Post reported last September that Anthropic was staring down a possible clash with the White House over concerns the AI firm carries a leftist bias.
Anthropic CEO Dario Amodei is a prominent Democratic donor, and Ford Foundation – a notorious left-leaning group – bought $5 million worth of Anthropic shares in March 2024.
“The Department of War’s relationship with Anthropic is being reviewed. Our nation requires that our partners be willing to help our warfighters win in any fight,” Pentagon spokesman Sean Parnell told The Post.
“Ultimately, this is about our troops and the safety of the American people.”
An Anthropic spokesperson noted that Claude is the first and only AI model currently used in the military’s classified systems. It was recently used in the US operation to capture Venezuelan dictator Nicolás Maduro, according to several reports.
“We are having productive conversations, in good faith, with DoW on how to continue that work and get these complex issues right,” the spokesperson told The Post.
Critics say they are concerned that unbridled access to AI will strengthen the Pentagon’s ability to target civilians.
Hegseth’s Department of War has argued that the exceptions Anthropic is seeking are too restrictive, insisting that the Pentagon should be allowed to use AI tools for “all lawful purposes.”
If Hegseth labels Anthropic a supply chain risk, the company could see some partners unwinding their business deals in order to maintain a good relationship with the Pentagon.
The Pentagon’s contract with Anthropic is valued at roughly $200 million – though this is just a fraction of the company’s $14 billion in annual revenue.
While OpenAI’s ChatGPT has seen more success among consumers, Anthropic has soared ahead in business deals. The company recently said it has partnerships with eight of the 10 largest US companies.
It also wouldn’t be a simple switch for the Pentagon, since rival AI bots from OpenAI, Google and xAI “are just behind” when it comes to government applications, a senior Trump administration official told Axios.
OpenAI, Google and xAI have all agreed to remove their chatbot safeguards for use in the military’s unclassified systems, but none of these bots are currently used in classified systems.
A senior administration official said the Pentagon is confident these three AI companies will agree to the looser standards, though another source told Axios that discussions are still ongoing.
