The Department of Defense has informed Anthropic’s leadership that the company and its products are now considered a supply chain risk, according to Bloomberg, which referenced a senior department official.
This designation follows weeks of conflict between the AI lab and the DOD. Anthropic CEO Dario Amodei has refused to permit military use of its AI systems for large-scale American surveillance or for autonomous weapons without human participation in targeting or firing decisions. The Department insists that AI usage should not be restricted by private entities.
Supply chain risk designations are usually applied to foreign threats. This designation mandates any entity working with the Pentagon to confirm they don’t use Anthropic’s models.
The Pentagon’s decision could disrupt both the company and its operations. Anthropic is the only frontier AI lab with systems ready for classified use. The U.S. military depends on Claude for its Iran campaign, using AI tools for efficient data management. Claude is integrated into Palantir’s Maven Smart System, essential for military operations in the Middle East, reports Bloomberg.
Critics regard the supply chain risk label as unprecedented. Dean Ball, a former Trump White House AI advisor, described it as a “death rattle” for the American republic, denouncing the government for abandoning strategic clarity for “thuggish” tribalism that treats domestic innovators worse than foreign adversaries.
Hundreds of OpenAI and Google employees have urged the DOD to rescind the designation and have called on Congress to challenge what might be an inappropriate action against an American tech company. They’ve also encouraged leaders to unite in refusing the DOD’s demands concerning AI model use for domestic surveillance and autonomous weaponry without human control.
TechCrunch contacted Anthropic for comments.
Amid this conflict, OpenAI has made an agreement with the Department to allow military use of its AI systems for “all lawful purposes.” Some employees expressed worries about the deal’s vague wording, which might permit uses Anthropic sought to prevent.
Amodei described the DOD’s actions as “retaliatory and punitive,” and reportedly linked the issue to his refusal to endorse or donate to President Trump. OpenAI President Greg Brockman has been a strong Trump supporter, recently donating $25 million to the MAGA Inc. Super PAC.