Anthropic Draws Line with Pentagon

The standoff escalated when Defense Secretary Pete Hegseth gave Anthropic a deadline to permit "all lawful uses" of its Claude AI model or be blacklisted. The Pentagon's position is that a private company's usage policy should not constrain military operations that are legal under U.S. law, similar to how contractors like Lockheed Martin operate. In response, the Trump administration ordered all federal agencies to stop using Anthropic's technology and designated the company a "supply chain risk." This designation, typically used for foreign adversaries, prohibits military contractors from doing any commercial activity with Anthropic, jeopardizing its wider enterprise business. The dispute puts a contract worth up to $200 million at risk; Google, OpenAI, and xAI received similar deals. Before the fallout, Anthropic was the only one of these firms with its model running on the Pentagon's classified networks, partly through a partnership with software analytics firm Palantir. CEO Dario Amodei has clarified the company's position is not a wholesale rejection of defense work but a refusal to permit two specific applications. He argues current AI is not reliable enough for autonomous weapons that remove humans from the loop and has offered to work with the DoD on R&D to improve reliability, an offer that was not accepted. The second red line is mass domestic surveillance, which Amodei says poses "serious, novel risks to our fundamental liberties." He notes that the law has not kept up with AI's ability to aggregate publicly available data into detailed profiles without a warrant, a practice that has drawn bipartisan concern in Congress. This conflict may benefit Anthropic's rivals. OpenAI previously removed its own ban on military applications, and officials say Elon Musk's xAI has agreed to the Pentagon's "all lawful purposes" policy for classified work. Musk sided with the administration, stating on X that "Anthropic hates Western Civilization."