Pentagon designates key AI contractor a ‘national security risk’
US Secretary of War Pete Hegseth has formally designated artificial intelligence firm Anthropic a “supply-chain risk to national security,” following President Donald Trump’s order to blacklist the company across the federal government.
Hegseth announced the decision Friday, accusing the company of attempting to impose ideological constraints on US military operations after Anthropic refused to remove safeguards limiting the use of its artificial intelligence systems for mass surveillance and fully autonomous weapons.
“Their true objective is unmistakable: to seize veto power over the operational decisions of the United States military,” Hegseth said in a statement posted on social media. “That is unacceptable... The Department of War must have full, unrestricted access to Anthropic’s models for every lawful purpose in defense of the Republic.”
“I am directing the Department of War to designate Anthropic a Supply-Chain Risk to National Security... This decision is final,” Hegseth stated.
The designation immediately bars any contractor, supplier, or partner working with the US military from engaging in commercial activity with Anthropic, significantly widening the impact beyond the Pentagon itself. Defense officials said the company will be permitted to continue providing services for up to six months to allow military systems to transition to alternative providers.
The move follows Trump’s sweeping directive ordering federal agencies to halt use of Anthropic technology after the company resisted Pentagon demands to lift contractual limits governing how its flagship AI model, Claude, could be deployed.
Anthropic has maintained two non-negotiable restrictions: prohibitions on mass domestic surveillance and on fully autonomous weapons operating without meaningful human control. Pentagon officials argue such safeguards could hinder operational flexibility during military crises.
The “supply-chain risk” designation – typically reserved for companies linked to foreign adversaries – marks an extraordinary step against a US-based technology firm deeply embedded in national security operations.
Anthropic was the first and only commercial AI developer to deploy advanced language models on classified Pentagon networks under a contract valued at up to $200 million. Its systems have supported intelligence analysis and military operational planning, including the raid that targeted Venezuelan President Nicolas Maduro.
