Anthropic confronted the Defense Department over contract restrictions after Defense Secretary Pete Hegseth demanded the company remove limits preventing military use of its models for certain operations. The Pentagon warned it could rescind a $200 million agreement and label Anthropic a ‘‘supply‑chain risk’’ if the company refused to broaden permitted uses. Anthropic says it will continue ‘‘good‑faith conversations’’ and support national security functions that align with its safety principles, while the administration accuses some AI firms of imposing ‘‘ideological constraints’’ that impede defense applications. The dispute unfolded amid broader federal contracting of AI firms and competing approvals for other companies to supply GenAI tools to the military. The standoff has implications for university researchers and labs that partner with commercial AI vendors and defense contractors: federal procurement decisions and export controls could accelerate or curtail academic access to advanced models, reshape research collaborations, and force institutions to reassess compliance and national‑security risk.