One Government, Two Voices: Recent Anthropic vs. Pentagon Developments Add to the Record Before the D.C. Circuit

In this new entry for Jones Walker’s AI Law & Policy Navigator, authors Andrew R. Lee & Michelle Ramsden discuss recent developments in the high-profile case Anthropic PBC v. U.S. Department of War, arguing that several events occurring shortly before oral arguments have significantly altered the factual record before the court. These include a White House meeting involving Anthropic CEO Dario Amodei, critical public remarks by Defense Secretary Pete Hegseth, and the Department of War’s announcement of AI contracts with seven major technology providers that excluded Anthropic.

The authors suggest these developments could affect the court’s assessment of both the government’s claimed need for Anthropic’s exclusion and the company’s First Amendment arguments. The Pentagon’s new partnerships with alternative AI vendors may weaken claims that Anthropic’s designation as a supply-chain risk is essential to national security, while Hegseth’s public criticism of Anthropic’s leadership could become relevant evidence in evaluating allegations of retaliation against the company for its views and policies.

The article also explores broader questions raised by the Department’s “AI-first” military strategy. With multiple frontier AI systems now being deployed on classified networks, the focus shifts from procurement decisions to the safety architectures governing these models. The authors note that questions about post-deployment control, operational safeguards, and government oversight apply equally to all contracted AI providers and may become increasingly important as military AI adoption expands.

Finally, the authors examine whether Anthropic’s refusal to relax certain usage restrictions represents an isolated business decision or an early warning sign of broader industry tensions between AI safety and government deployment demands. They argue that future court proceedings and related appeals will likely help determine whether Anthropic was an outlier or a “canary in the coal mine,” highlighting challenges that could eventually affect the entire AI industry as governments seek greater access to advanced AI systems.