Defense Secretary Pete Hegseth has given Anthropic until Friday evening to provide the military with broad access to its artificial intelligence models, according to sources familiar with the discussions.
If Anthropic fails to comply, Hegseth has threatened to designate the company a “supply chain risk”, a label typically reserved for foreign adversaries that would require Defense Department vendors to certify they do not use Anthropic’s technology, or potentially invoke the Defense Production Act, which allows presidential control over domestic industries in the interest of national security.
Negotiations between Anthropic and the Department of Defense have stalled over usage terms. The company is seeking assurances that its models will not be used for autonomous weapons or mass surveillance of Americans, while the DoD insists on access for “all lawful use cases” without limitation. The deadline was reportedly set during a Tuesday morning meeting between Hegseth and Anthropic CEO Dario Amodei.
Anthropic, founded in 2021 by former OpenAI researchers, is best known for its Claude family of AI models. It secured a $200 million DoD contract last year and, until recently, was the only AI company authorized to deploy models on classified Pentagon networks. xAI has since become the second firm cleared for classified deployment.
Despite tensions with the Trump administration, including public criticism from White House AI and crypto czar David Sacks, Anthropic maintains a broad enterprise base, with more than 500 customers spending over $1 million annually. The company also recently closed a $30 billion funding round at a $380 billion valuation.
