Anthropic to contest Pentagon's supply chain designation in court

Dario Amodei stated on Thursday that Anthropic intends to legally challenge the Defense Department's decision to designate the AI company as a supply chain risk, a classification he has described as "legally unsound."

This statement follows the Department's official designation of Anthropic as a supply chain risk, concluding a weeks-long dispute regarding the appropriate level of military control over AI systems. Such a designation can prohibit a company from engaging with the Pentagon and its contractors. Amodei firmly asserted that Anthropic's AI should not be utilized for mass surveillance of American citizens or for fully autonomous weaponry. However, the Pentagon maintained it required unrestricted access for "all lawful purposes."

In his remarks, Amodei clarified that the vast majority of Anthropic's clientele remain unaffected by this supply chain risk designation.

"Regarding our customers, it clearly applies solely to the use of Claude as a direct component of contracts with the Department of War, not to all usage of Claude by customers who hold such contracts," he explained.

Offering a preview of Anthropic's likely legal argument, Amodei characterized the Department's letter imposing the supply chain risk label as limited in scope.

"Its purpose is to safeguard the government rather than penalize a supplier; indeed, the law mandates that the Secretary of War employ the least restrictive means necessary to achieve the objective of securing the supply chain," Amodei stated. "Even for Department of War contractors, the supply chain risk designation does not (and cannot) restrict uses of Claude or business dealings with Anthropic if those are unrelated to their specific Department of War contracts."

Amodei reiterated that Anthropic had been engaged in productive discussions with the Department over recent days. These talks are believed by some to have been disrupted following the leak of an internal memo he circulated to staff. In that document, Amodei characterized rival OpenAI's arrangements with the Department of Defense as "safety theater."

OpenAI has subsequently entered into an agreement to work with the Defense Department in Anthropic's stead, a move that has provoked internal backlash among OpenAI employees.

In his Thursday statement, Amodei apologized for the memo's leak, asserting the company did not intentionally disclose it nor instruct anyone else to do so. "It is not in our interest to escalate the situation," he said.

Amodei noted the memo was drafted within "a few hours" of a series of rapid developments: a presidential Truth Social post announcing Anthropic's removal from federal systems, followed by Defense Secretary Hegseth's supply chain risk designation, and finally the Pentagon's partnership announcement with OpenAI. He apologized for its tone, calling that "a difficult day for the company," and stated the memo did not reflect his "careful or considered views." Written six days prior, he added, it now represents an "out-of-date assessment."

He concluded by emphasizing that Anthropic's foremost priority is to ensure American military personnel and national security professionals retain access to critical tools amidst ongoing major combat operations. Anthropic is currently supporting certain U.S. operations in Iran, and Amodei committed that the company would continue providing its models to the Defense Department at "nominal cost" for "as long as necessary to facilitate that transition."

Anthropic could contest the designation in federal court, likely in Washington, D.C. However, the legal framework underpinning the decision presents a significant challenge for appeals, as it restricts conventional avenues for challenging government procurement actions and grants the Pentagon considerable latitude on national security grounds.

Or as Dean Ball—a former Trump administration White House advisor on AI who has criticized Hegseth's handling of Anthropic—observed: "Courts are generally quite hesitant to second-guess the government on what constitutes a national security issue... The legal threshold to successfully challenge such a determination is exceptionally high. But it is not an impossibility."