Anthropic said Thursday that the Pentagon’s designation of AI companies as a “supply chain risk to U.S. national security” will not have as big an impact on its business partners as Defense Secretary Pete Hegseth suggested last week.
Hegseth announced in a post on X on Friday that “no contractor, supplier, or partner doing business with the U.S. military may conduct commercial activities with Anthropic” as part of the supply chain risk designation as a result of contract negotiations with the AI company breaking down. This is a label typically given to companies with hostile foreign relations.
But Anthropic CEO Dario Amodei said in a statement Thursday that the language in the Pentagon’s letter regarding the designation means military contractors will be prohibited from using the company’s AI model, Claude, only “as a direct part.” It is part of a contract with the Department of the Army and not all Claude use by customers with such contracts. ”
The Pentagon and Anthropic are at odds over restrictions the company has placed on the use of Claude, the first AI system used in classified military networks. The government’s response raised concerns that the blacklist would cut into Anthropic’s business.
But Anthropic said relevant supply chain laws limit the scope of the Department of Defense designation.
“Even if you are a Department of the Army contractor, the supply chain risk designation does not (and cannot) limit your use of Claude or your business relationship with Anthropic, if unrelated to your specific Department of the Army contract,” Amodei wrote.
Microsoft, one of Anthropic’s biggest partners, agreed. A spokesperson told CNN: “Our attorneys have reviewed this designation and have concluded that man-made products, including Claude, will continue to be available to our customers outside the Department of the Army.” .– We look forward to continuing to work with Anthropic on non-defense projects through platforms such as M365, GitHub, and Microsoft’s AI Foundry. ”
Google and Amazon issued similar statements, saying Anthropic products will continue to be available through their platforms for all non-Department of Defense operations.
A senior Pentagon official confirmed to CNN that the department has formally notified human leadership of the supply chain risk designation.
“From the beginning, one fundamental principle has been that the military can use technology for any lawful purpose. The military will not allow vendors to enter the chain of command and endanger our warfighters by restricting the lawful use of critical capabilities,” the official said.
“We do not believe this action is legally sound and see no other option than to challenge it in court,” Amodei wrote on Thursday.
The Pentagon wanted a contract change that would give the department “all lawful uses” of Claude. However, negotiations stalled over Anthropic’s two red lines. Concerns about AI being used in autonomous weapons and concerns about AI being used for mass surveillance of American citizens. The Pentagon said this red line was unacceptable.
Amodei said last week that he could not “in good conscience” comply with the Pentagon’s request. Shortly after Hegseth declared Anthropic at supply chain risk, President Donald Trump directed all federal agencies to stop using Anthropic’s products within six months.
Mr. Amodei indicated that negotiations have continued between the two sides in recent days despite divided public opinion.
“I would like to reiterate that over the past several days, we have had productive conversations with the Department of the Army about how we can serve the Department of the Army adhering to two narrow exceptions and how to ensure a smooth transition when that is not possible,” Amodei said.
Undersecretary of Defense Emile Michael reacted, writing: “We would like to end all speculation. There are no active @DeptofWar negotiations with @AnthropicAI,” Amodei said on X after his statement was published.
Mr. Amodei also apologized for the contents of a leaked internal memo sent to staff on Friday and published by the Information newspaper, in which he said one of the real causes of the controversy was “that we do not praise Mr. Trump in a dictator-like manner.”
“It was a difficult day for the company. I apologize for the poor quality of my posts,” Amodei wrote. “This does not reflect my careful or considered views. It was also written six days ago and is an outdated assessment of the current situation.”
Sam Altman, CEO of rival AI company OpenAI, has told the Pentagon publicly and privately that Anthropic should not be classified as a supply chain risk. (OpenAI last week signed its own contract with the Department of Defense for the use of AI models on classified systems, claiming it respects the same red lines set by humanity.)
On Thursday, 30 former military and intelligence officials and technology policy leaders sent a joint letter to Congress urging Congress to investigate the “dangerous precedent” set by the Pentagon’s Anthropic attack.
According to Reuters, major tech industry groups in Washington sent a letter to the administration warning of the negative effects of designating U.S. companies as supply chain risks.
Members of Congress have also criticized the administration. Republican Sen. Thom Tillis told Axios that public conflict between the Pentagon and humanity is “secondary.” Democratic Sen. Kirsten Gillibrand said in a statement Thursday that supply chain risk designations are “shortsighted, self-defeating, and a gift to our adversaries.”
CNN’s Haley Britsky contributed reporting.
