According to people familiar with the matter. The Pentagon issued a warning to Anthropic PBC that it will terminate its military contract with the artificial intelligence startup on Friday if it fails to meet the terms of the government's use of its technology. During a meeting between CEO Dario Amodei and Defense Secretary Pete Hegseth on Tuesday, U.S. officials threatened to list Anthropic as a supply chain risk or invoke the Defense Production Act to force use of the company's artificial intelligence software, these people said.
The ultimatum marks an escalating dispute between the Department of Defense and the artificial intelligence startup, with Anthropic insisting on setting guardrails on the use of its Claude AI tools. If the Pentagon follows through on its threat, projects of up to $200 million that Anthropic has previously agreed to carry out for the military will be at risk.
One of the people said Amodei outlined Anthropic's conditions at the meeting: The U.S. military would not be able to use its products to autonomously target enemy combatants or conduct mass surveillance of U.S. citizens. Amodei emphasized that these situations have not yet materialized in actual operations, the person said.
Anthropic said in a statement after the meeting: "We continue to have honest conversations about usage policies to ensure that Anthropic can continue to support the government's national security mission within the capabilities of the model's reliability and responsibility."
Anthropic has no plans to relax restrictions on military use, people familiar with the matter said.