Anthropic to challenge DOD’s supply-chain label in court

Dario Amodei said Thursday that Anthropic plans to challenge in court the Defense Department’s decision to label the AI ​​company as a supply chain risk, a designation it has called “legally unsound.”

The statement comes just hours after the DOD officially designated Anthropic as a supply chain risk following a weeks-long dispute over how much control the military should have over AI systems. Identifying a supply chain risk can deter a company from working with the Pentagon and its contractors. Amodei drew a firm line that Anthropic’s AI will not be used for mass surveillance of Americans or for fully autonomous weapons, but the Pentagon opined that it should have unrestricted access for “all lawful purposes.”

In its statement, Amodei said the vast majority of Anthropic’s customers are unaffected by the designation of a risk in the supply chain.

“As far as our customers are concerned, this clearly only applies to customer use of Claude as a direct part of contracts with the War Department, and not all uses of Claude by customers who have such contracts,” he said.

In a preview of what Anthropic will likely argue in court, Amodei said the ministry’s letter labeling the company as a supply chain risk is limited in scope.

“It is there to protect the government, not to punish a supplier; in fact, the law requires the Secretary of War to use the law least restrictive means required to achieve the goal of protecting the supply chain,” Amodei said. “Even for Department of War contractors, the designation of supply chain risk does not (and cannot) limit the use of Claude or business relationships with Anthropic if not related to their specific Department of War contracts.”

Amodei reiterated that Anthropic had had productive conversations with the Defense Department in recent days, conversations that some suspect were derailed when an internal memo he sent to staff was leaked. In it, Amodei characterized rival OpenAI’s dealings with the Ministry of Defense as “security theater.”

OpenAI has signed a deal to partner with the DOD in Anthropic’s place, a move that has sparked backlash among OpenAI staff.

Amodei apologized for the leak in its statement Thursday, claiming that the company did not intentionally share the memo or direct anyone else to do so. “It is not in our interest to escalate the situation,” he said.

Amodei said the memo was written within “a few hours” of a series of announcements, including a presidential Truth Social post saying Anthropic would be removed from federal systems, then Defense Secretary Pete Hegseth’s guidance on supply chain risks, and finally the announcement of the Pentagon’s deal with OpenAI. He apologized for the tone, calling it “a difficult day for the company” and saying the memo did not reflect his “prudent or considered views.” It was written six days ago, he added, and is now an “outdated review.”

He concluded by saying that Anthropic’s top priority is to ensure that U.S. soldiers and national security experts continue to have access to important tools amid ongoing major combat operations. Anthropic is currently supporting some U.S. operations in Iran, and Amodei said the company would continue to supply its models to the Defense Department at “nominal cost” for “as long as it takes to make that transition.”

Anthropic could challenge the designation in federal court, likely in Washington, but the law behind the decision makes it more difficult to challenge because it limits the usual ways companies can challenge government procurement decisions and gives the Pentagon broad discretion on national security issues.

As Dean Ball — a former Trump-era White House AI adviser who has spoken out against Hegseth’s handling of Anthropic — put it: “Courts are quite reluctant to second-guess the government on what is and isn’t a national security issue… There’s a very high bar you have to cross to do that. But it’s not impossible.”