Anthropic in "legal limbo": two US courts can't agree on whether the military can use Claude

Anthropic finds itself in a difficult legal position: a US appeals court has issued a ruling that directly contradicts a lower court's decision from March. At the heart of the dispute is whether the US military can use the Claude model, and if so, to what extent. Until the conflict between the courts is resolved, the company is effectively stuck in a zone of legal uncertainty, which it describes itself as a "supply-chain risk."

The legal collision comes down to two court rulings that don't line up. The lower court, which issued its ruling in March, took a specific position on how and under what terms Anthropic may license its technology to federal agencies. The appeals court takes a different view.

When two courts disagree on the same legal question, neither side — neither the developer nor the customer — can confidently rely on an agreement that the next court up may revisit. That is exactly where Anthropic now stands. The timing is especially sensitive given the company's ambitions.

Claude ranks among the leading language models, competing with OpenAI's GPT-4o and Google's Gemini. Anthropic has aggressively pushed its technology into the government and enterprise sectors, with Amazon as its key ally: Amazon has invested billions of dollars in the company and integrated Claude into its AWS cloud platform. Any legal cloud over its ability to work with military customers hits revenue and negotiating leverage directly.

Anthropic is valued at tens of billions of dollars, and a significant share of that figure rests specifically on growth prospects in the public sector. The term "supply-chain risk," which Anthropic itself uses, points to a broader problem than it might first appear. In software and AI, the "supply chain" is a complex structure of dependencies: the base model, the API, the cloud provider, partner integrations, and end-user applications.

If the legal status of the base model itself is in question, the instability spreads down the entire chain — from integrator startups to major defense contractors that have already built Claude into their workflows. Over the past two years, the US military and intelligence agencies have moved from cautious AI experiments to full-scale deployment of language models in intelligence operations, logistics, and strategic planning. The biggest players in the market — Microsoft, Google, Oracle, Palantir — are competing for defense AI contracts.

Anthropic, which positions itself as a company focused on safe and predictable AI, looked like a particularly attractive partner for agencies that cannot afford unpredictable systems. Now military customers have legal grounds to pause and rethink how heavily they depend on Claude. Conflicts between courts are usually resolved by a higher authority — either by taking the case to the Supreme Court or by establishing new precedent within the appellate circuit.

Until then, Anthropic has to operate under uncertainty: signing contracts while knowing their legal basis may be revisited. For the industry as a whole, this is a telling test case. OpenAI, Google, and other contenders for government AI contracts are watching how it plays out: the legal framework the courts ultimately settle on will either open the door to broad AI adoption in the defense sector or create constraints the entire industry will have to reckon with.