Even though the Pentagon has recently classified Anthropic as a supply-chain risk, the company is still engaging with senior figures from the Trump administration.

Indications of a warming relationship — or at least a sense that not all factions of the administration aimed to sever ties with Anthropic — surfaced previously, with reports indicating that Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell encouraged leaders of major banks to experiment with Anthropic’s new Mythos model.

Jack Clark, co-founder of Anthropic, appeared to affirm this, stating that the ongoing dispute regarding the supply-chain risk label is a “narrow contracting issue” that would not hinder the company’s readiness to inform the government about its newest models.

On Friday, Axios disclosed that Bessent and White House Chief of Staff Susie Wiles had a meeting with Anthropic CEO Dario Amodei. The White House characterized this as an “introductory meeting” that was “productive and constructive.”

“We talked about collaboration opportunities, as well as common approaches and protocols to tackle the challenges related to scaling this technology,” the White House stated.

In a similar vein, Anthropic released a statement confirming that Amodei had engaged in a “productive discussion with senior administration officials on how Anthropic and the U.S. government can collaborate on crucial shared priorities such as cybersecurity, maintaining America’s leadership in the AI race, and AI safety.”

The company expressed that it’s “looking forward to continuing these conversations.”

The conflict between Anthropic and the Pentagon appears to have initiated after unsuccessful discussions regarding the military’s application of Anthropic’s models; the AI firm aimed to preserve safeguards surrounding the use of its technology for fully autonomous weaponry and extensive domestic surveillance. (OpenAI swiftly revealed a military agreement of its own, triggering some consumer backlash.)

In turn, the Pentagon labeled Anthropic a supply-chain risk — a designation typically reserved for foreign adversaries that could severely restrict the government’s utilization of Anthropic’s models. The company is contesting that label in court. 

Nonetheless, it seems the remainder of the Trump administration does not share the Pentagon’s antagonism, with an administration source informing Axios that “every agency” except the Department of Defense is interested in using the company’s technology.