Anthropic challenges Department of War designation as AI dispute escalates

The Department of Defense, known under the Trump administration as the Department of War, has just officially designated AI company Anthropic a "supply-chain risk" to national security. But Anthropic isn't buying it.

"We do not believe this action is legally sound, and we see no choice but to challenge it in court," Anthropic CEO Dario Amodei wrote in a statement responding to the ongoing dispute. Amodei also emphasized that the designation does not affect the majority of Anthropic customers.

Anthropic is "proud of the work" it has done alongside the federal government in "supporting frontline warfighters with applications such as intelligence analysis, modeling and simulation, operational planning, cyber operations, and more," Amodei said.

The dispute began over the potential use of Anthropic's AI technology to carry out mass domestic surveillance, and to power autonomous weapons like drones. The relationship between Anthropic and the U.S. military deteriorated last week, after the AI company won a $200 million contact from the federal government — but sought guarantees that its technology would not be used for surveillance, or weapons that can fire without humans in the loop.  

The U.S. government would not agree to Anthropic's terms and threatened to designate the company a supply-chain risk, which it has now done. Trump also issued an executive order that tells every federal agency to stop using Anthropic's AI.

Amodei apologized for the leak of an internal memo, and confirmed recent reports that Anthropic and the Department of Defense have re-entered negotiations. If the two sides can't come to terms, Anthropic will assist the government in a transitional period, Amodei said.

The CEO also referenced the U.S. government's deal with OpenAI, made in the wake of the Anthropic dispute. Even OpenAI "characterized" its deal with the U.S. government as "confusing," he said.

OpenAI CEO Sam Altman was forced to address the deal after receiving significant blowback from users.

"Our most important priority right now is making sure that our warfighters and national security experts are not deprived of important tools in the middle of major combat operations," Amodei wrote. "Anthropic will provide our models to the Department of War and national security community, at nominal cost and with continuing support from our engineers, for as long as is necessary to make that transition, and for as long as we are permitted to do so."

As the Wall Street Journal previously reported, the U.S. military has already used Anthropic's Claude models to aid in carrying out strikes in Iran.