The Pentagon has asked Boeing and Lockheed Martin to assess their reliance on Anthropic’s AI model Claude, in what appears to be a first step toward potentially designating Anthropic as a supply chain risk. That designation is typically reserved for companies from adversarial nations such as Chinese tech giant Huawei, making its potential use against a leading American AI company unprecedented.
The move follows a tense meeting on Tuesday in which Defense Secretary Pete Hegseth gave Anthropic CEO Dario Amodei a deadline of 5:01 p.m. Friday to agree to the Pentagon’s terms. If Anthropic does not comply, Hegseth warned the administration would either invoke the Defense Production Act to compel Anthropic to modify Claude for military use, or declare the company a supply chain risk. Wednesday’s outreach to major contractors suggests the Pentagon may be leaning toward the supply chain risk route.
The core dispute is over Anthropic’s refusal to lift its safety guardrails. The Pentagon wants to use Claude for all lawful purposes without having to seek approval for individual use cases. Anthropic has held firm on blocking Claude’s use for mass surveillance of Americans and for autonomous weapons systems that fire without human involvement. CEO Dario Amodei has personally highlighted both areas as among the most significant dangers posed by AI.
The stakes are significant. Claude is currently the only AI model operating in the military’s classified systems and was used during the operation to capture former Venezuelan President Nicolás Maduro through Anthropic’s partnership with Palantir. The Pentagon acknowledges Claude is the most capable model for a number of military use cases but is exploring alternatives. Elon Musk’s xAI has already signed a deal to enter classified systems under the all-lawful-use standard Anthropic has rejected. Google’s Gemini is described as a strong alternative, and both Google and OpenAI are in negotiations about moving into classified systems, though the Pentagon says they would also need to lift their own safeguards to do so.
A supply chain risk designation could be a significant blow to Anthropic if government-adjacent companies begin removing Claude from their operations. However, some observers note the Pentagon’s outreach to contractors may be more of a pressure tactic than an immediate move to cut ties. Anthropic described Tuesday’s meeting as a continuation of good-faith conversations about usage policy. The Friday deadline leaves little time for a resolution.
