Anthropic is at risk of losing significant U.S. government business. Previously, the U.S. Department of Defense stated that it constituted a "supply chain risk." Such penalties have never been imposed on a U.S. company, let alone a cutting-edge new technology company in a government priority area, according to contracting and national security experts. They said the Pentagon's move could set a dangerous precedent for companies seeking to innovate like Anthropic has in artificial intelligence.
“Taking this tool against a domestic AI company sends a worrying signal that could stifle innovation and weaken the technology ecosystem the United States needs to remain competitive,” said Morgan Plummer, vice president of policy at Americans for Responsible Innovation. “These powers are intended to prevent foreign adversaries from entering supply chains, not to punish U.S. companies for building security safeguards into their technology.”
The Pentagon has officially notified Anthropic PBC that it has determined that the company and its products pose a risk to the U.S. supply chain.
The decision could jeopardize Anthropic's $200 million Pentagon contract for classified artificial intelligence tools and could prevent it from partnering with other companies on defense business. While the amount is only a fraction of the company's $20 billion in expected 2026 revenue, the "supply chain risk" label could cast a long-term shadow on a company whose AI tools are quickly gaining popularity in the corporate world.
The negotiations followed weeks of negotiations, but they broke down last week after Anthropic demanded assurances that its artificial intelligence technology would not be used to mass-surveil Americans or deploy autonomous weapons. That prompted Trump to order federal agencies to stop working with Anthropic, and Defense Secretary Hegseth to threaten rare supply chain exclusions.
The Pentagon is relying on Section 3252 of the U.S. Armed Forces Act, which allows companies to be banned from becoming contractors if they are determined to endanger the security of the supply chain. The terms define risk as the possibility that “an adversary may sabotage, maliciously implant adverse functionality, or otherwise subvert” the technology or services provided.
The provision requires the U.S. Secretary of Defense to follow established procedures, including demonstrating supply chain risks and demonstrating that no other less intrusive measures are available. Hegseth has sent letters to top Republicans and Democrats on the House and Senate Armed Services, Appropriations and Intelligence committees informing Congress of his decision, according to the letter seen.
"This decision is based in part on the War Department's risk analysis and the opinion of senior War Department personnel that the covered entities' restrictions on the use of their products and services pose a national security risk to the War Department's supply chain," Hegseth wrote, referring to Anthropic.
"We believe this action is legally untenable and we have no choice but to go to court," Anthropic CEO Dario Amodei said in a blog post Thursday.
Alan Rozenshtein, a professor at the University of Minnesota Law School, said supply chain risk determinations under Section 3252 should not apply to Anthropic’s situation at all. He pointed out that when Congress enacted the law, it was intended to crack down on foreign companies "using malware, backdoors or sabotage methods to invade government systems," rather than targeting American companies like Anthropic.
"If this counts as a supply chain risk, then every time the government disagrees with an American company on the terms of any contract, it can designate that company as a supply chain risk and destroy it. I don't think that's constitutional," Rozenshtein said.