'Incoherent' Pentagon Pete's Anthropic AI Ultimatum Deemed 'Insane' - The Daily Beast

The defense secretary’s desperate ploy to make a top Pentagon contractor cave to his demands is raising red flags for experts.

“It doesn’t make any sense,” Dean Ball, a former AI adviser in the Trump administration who helped craft the president’s AI plans, told Politico about threats that Defense Secretary Pete Hegseth made to Anthropic CEO Dario Amodei, 43, on Tuesday regarding contracts the AI company has with the agency.

During the Tuesday meeting between Hegseth and Amodei, Hegseth, 45, gave the CEO of the AI safety and research company an ultimatum: give the military unfettered access to its AI model, or be considered a “supply chain risk” and have ties with the Pentagon cut.

The meeting didn’t begin under promising circumstances. Prior to its start, one defense official described it as “a s--t-or-get-off-the-pot-meeting”.

Hegseth’s threat—to label Anthropic a “risk” and cut ties with the company, or to invoke the Defense Production Act (DPA), which would require the company to provide its technology on the Pentagon’s terms without its current ethical restrictions—has left experts on the matter scratching their heads.

“You’re telling everyone else who supplies to the DoD you cannot use Anthropic’s models, while also saying that the DoD must use Anthropic’s models,” Ball said about Hegseth’s ultimatum to Anthropic, labeling the move as “incoherent.”

Ball added that in presenting the two policies, it was “a whole different level of insane to move up and say we’re going to do both of those things.”

Amodei has so far resisted the push from the defense secretary, who prefers to be called the “secretary of war,” to grant unrestricted, unconstrained access to Anthropic’s AI technology, specifically the large language model Claude.

Tensions between Anthropic and Hegseth escalated amid reports that Claude was used by the Pentagon in the Trump administration’s invasion of Venezuela earlier this month and the capture of Venezuelan President Nicolás Maduro.

Axios reported that during the Tuesday meeting, Hegseth referenced the Pentagon’s claim that Anthropic had raised concerns with its partner Palantir over the use of Claude during the Venezuela mission—a claim Amodei denied.

Katie Sweeten, a technology lawyer and former Department of Justice official, told Politico that Hegseth’s threat is “contradictory.”

“I don’t know how you can both use the DPA to take over this product and also at the same time say this product is a massive national security risk,” Sweeten told Politico, warning that Hegseth’s “very aggressive” negotiating could be devastating for future partnerships with Silicon Valley.

The feud with Hegseth risks breaking a $200 million Pentagon contract with Anthropic, signed in July to develop “AI capabilities that advance U.S. national security.”

“The only reason we’re still talking to these people is we need them and we need them now. The problem for these guys is they are that good,” a Defense official told Axios about the AI company.

The Daily Beast has contacted the Defense Department for comment.