Anthropic finds itself in a contradictory position: its Claude models are actively being used in U.S. military targeting decisions during the ongoing conflict with Iran, while simultaneously being phased out by defense contractors following a Trump administration directive to designate the company as a supply-chain risk. The DoD was given six months to wind down Anthropic usage, but the U.S.-Israel attack on Tehran began before that process completed. Washington Post reporting reveals Claude is being used alongside Palantir's Maven system for real-time target suggestion and prioritization. Meanwhile, Lockheed Martin and other contractors are already replacing Claude, with a J2 Ventures partner noting 10 portfolio companies are actively switching away. The key unresolved question is whether Secretary Hegseth will formally execute the supply-chain risk designation, which would trigger legal battles.
Sort: