RismadarVoice Reporters
September 25, 2026
A United States federal appeals court has upheld the Pentagon’s decision to blacklist artificial intelligence company Anthropic from military contracts, rejecting the company’s challenge to its designation as a national security supply-chain risk.
The US Court of Appeals for the District of Columbia Circuit ruled 2-1 on Friday that the Pentagon had sufficient grounds for the designation after Anthropic refused to remove safeguards restricting the use of its Claude AI models for fully autonomous weapons and mass surveillance.
The dispute began earlier this year when the Pentagon pressed Anthropic to loosen restrictions governing military use of Claude.
Anthropic resisted the demand, maintaining that certain safeguards were necessary because of safety, privacy and ethical concerns surrounding advanced artificial intelligence.
The Pentagon subsequently designated Anthropic a national security supply-chain risk in March, restricting the company’s technology from Defence Department work.
In its ruling, the appeals court majority said the Pentagon’s concerns about relying on an AI system whose developer could restrict certain functions were reasonable.
The judges rejected Anthropic’s argument that the designation amounted to retaliation for the company’s position on AI safety and ethics.
The decision allows the Pentagon to continue removing Claude models from its systems and restricting their use in Defence Department-related work.
Anthropic said it disagreed with the judgment and was considering further legal action.
“We respectfully disagree with the court’s decision,” the company said, adding that it remained confident in its position and could seek further judicial review.
The latest ruling is separate from a case in California, where a federal judge previously found that a parallel government action against Anthropic was unlawful.
The appeals court acknowledged the separate decision but concluded that the Pentagon’s supply-chain risk designation at issue in the Washington case could stand.
The legal battle has become a major test of how much control AI developers can retain over the military use of their technology and how far the US government can go in restricting companies whose safeguards conflict with national-security requirements.
