A US appeals court has affirmed the Department of Defense’s decision to blacklist Anthropic’s technology, ruling that the Pentagon has the authority to exclude the AI developer for withholding certain features from military use.
What Happened
In a 2-1 ruling, the US Court of Appeals for the District of Columbia Circuit determined that the Trump administration acted within its rights under the Supply Chain Security Act and the Constitution. The court denied Anthropic’s petitions for review, noting that malicious intent was not a prerequisite for the blacklisting. This decision follows an earlier ruling in April where the same court denied Anthropic’s emergency motion for a stay.
Why It Matters
The ruling highlights the growing tension between AI safety constraints and military operational requirements. The judges described the case as raising "profoundly difficult questions about the appropriate military uses of an almost unimaginably powerful new technology." The court acknowledged the competing risks cited by both parties: the government’s concern that "overly constrained AI models" might shut down unexpectedly and cause military operations to fail, versus Anthropic’s warning that "unconstrained AI models" could "hallucinate inappropriate targets for lethal military force." Ultimately, the court stated that balancing these risks is the responsibility of Defense Secretary Pete Hegseth and the President, not the judiciary.
The Bottom Line
Anthropic remains blacklisted by the Pentagon for refusing to enable specific Claude features for military applications, with the court affirming that the executive branch has the discretion to determine how to balance AI safety constraints against operational reliability.