The partnership had been announced in July 2025 and integrated Claude into classified military networks, including the Pentagon's Maven Smart System for target processing. Months of negotiations between Amodei and Pentagon leadership, particularly Undersecretary Emil Michael, deteriorated into a public legal battle when Anthropic filed for an injunction against the supply chain risk designation in March 2026. The company has since resumed talks with the Pentagon but has not retreated from its original safety conditions. OpenAI secured a new Department of War contract after the Anthropic negotiations collapsed.
For practitioners, the case exposes a fundamental tension between commercial AI ethics standards and national security demands—one likely to recur across the defense contracting landscape. The "supply chain risk" designation sets a precedent for how the administration may pressure technology vendors on AI governance. Attorneys advising AI companies on government contracts should anticipate similar pressure campaigns and consider the enforceability and strategic costs of maintaining safety guardrails against executive pressure. The ongoing litigation will clarify whether such designations survive judicial review.