Legal Precedent: Anthropic vs. Pentagon

By March 7, 2026, the crisis surrounding Anthropic reached its climax. After formally being assigned the status of a "supply-chain risk" by the US Department of Defense, the startup officially announced its intention to sue the government.

The situation is unprecedented for the generative AI market. The Pentagon didn't just refuse the company's services; it triggered a chain reaction: the "supply chain threat" status means that no integrator working with the DoD (Palantir, Anduril, Lockheed Martin) is allowed to use the Claude API, even in auxiliary circuits. Anthropic's lawsuit is an attempt to save its B2B business. Simultaneously, the Pentagon is restructuring its AI division, appointing new leaders to strictly control vendors. The Zero Trust architecture is now applied not only to code but also to the political loyalty of the supplier.

Source: Reuters / DoD Press