Donald Trump boasted about severing ties between the US military and Anthropic on Thursday, the same day multiple reports said that negotiations between the Department of Defense and the AI startup had resumed.

They’re among the latest developments in the twisting rift between the US government and the AI company.

“Well, I fired Anthropic. Anthropic is in trouble because I fired [them] like dogs, because they shouldn’t have done that,” Trump told Politico on Thursday.

Hours later, the Pentagon officially designated Anthropic a “supply chain risk”, a move that prevents all government contractors from using the company’s technology. The label has never been used before against a US company.

“DoW officially informed Anthropic leadership the company and its products are deemed a supply chain risk, effective immediately,” a Pentagon official told Bloomberg.

But at the same time, there were reports of a thawing. Negotiations had restarted between the Pentagon and Anthropic over the military’s use of the company’s AI and the contract between the two, according to the Financial Times and Bloomberg. Anthropic’s products, which include the popular Claude chatbot and coding assistant, are integrated into Palantir’s Maven system, a newly vital tool of military intelligence that was used in recent strikes on Iran, according to the Washington Post.

Anthropic’s CEO, Dario Amodei, has been discussing the Pentagon’s contract with Emil Michael, the undersecretary of defense for research and engineering and a former Uber executive, per Bloomberg. The two strongly dislike one another, the New York Times reported.

Trump last week ordered the entire federal government to cease using Anthropic’s tech, after the company refused a deal with the government over concerns its model could be used for domestic mass surveillance or fully autonomous weapons.

Both the state and treasury departments already began severing ties, according to their respective heads. Pete Hegseth, the US defense secretary, verbally designated Anthropic a “supply chain risk” last week and Amodei has said his company will sue over the label.

On Thursday, Amodei sent a scorching message to his employees disparaging the Trump administration and rival OpenAI, which late last week jumped into the void left by the government and Anthropic’s breakup, and announced its own deal with the military. Amodei called Sam Altman, OpenAI’s CEO, “mendacious” and said Altman’s claims that the Pentagon would abide by safeguards amounted to nothing more than “safety theater”. From his end, Altman acceded to employees in an internal message this week that the company would have no control over how the military used OpenAI’s technology.

The retributive blacklisting of the company could cause significant financial harm. Anthropic’s most recent round of financing, about $60bn, is in jeopardy over the fracas with the Pentagon, per Axios. Silicon Valley, including OpenAI, has backed Anthropic in the fight over the designation as a supply chain risk.

As the defense secretary vowed punitive measures against Anthropic, his agency also announced a deal with OpenAI to use its technology for military operations in its classified network. The timing, as well as questions over ethics, elicited backlash that propelled Claude’s app to the top of the download charts in the US. In the following days, Altman said he would amend the agreement with the department and admitted his company’s conduct appeared “opportunistic and sloppy”.