A divided judicial panel found on Friday that Anthropic's internal AI safety guardrails pose a supply-chain threat to U.S. military operations. The D.C. Circuit Court of Appeals ruled 2-1 that the Pentagon had "ample support" to blacklist the company and its Claude AI from government contracts, specifically noting that the model's encoded restrictions had caused uncertainty during the Jan 3 military operation to capture Nicolás Maduro. The court determined that an AI provider's ability to refuse military tasks requested by government users constitutes a statutorily covered national-security risk.

Anthropic argues the designation has already cost it billions in business and damaged its reputation ahead of an expected IPO. While this ruling is a win for the Defense Department, a separate federal ruling in California by Judge Rita Lin still blocks a broader government-wide prohibition under a different statute. Anthropic disagrees with the D.C. Circuit's decision and is considering further review, while the ruling likely increases the Pentagon's leverage in contract negotiations with other AI suppliers including OpenAI, Google, and xAI.