Work Templates
Services
AI News About Contact
Talk to an engineer Sign In
Policy

DC Circuit Upholds Pentagon Blacklist of Anthropic Over Feature Refusal

A federal appeals court affirmed the Defense Department’s right to bar Anthropic after the AI firm refused to enable Claude capabilities for the military, signaling tighter supply‑chain control for contractors.

BuildPreview AI Desk 3 min read
Cover for: DC Circuit Upholds Pentagon Blacklist of Anthropic Over Feature Refusal
Photo: Wesley Tingey via Unsplash · Story via arstechnica.com

The U.S. Court of Appeals for the District of Columbia Circuit voted 2‑1 to uphold the Department of Defense’s decision to blacklist Anthropic after the company declined to enable certain Claude features for military use, a move that keeps the AI firm off federal procurement lists.

Key points

  • The panel affirmed the Pentagon’s authority under 41 U.S.C. § 4713 to designate Anthropic a supply‑chain risk.
  • Judges Gregory Katsas and Neomi Rao, both Trump appointees, authored the majority opinion.
  • Anthropic’s emergency stay request was denied earlier this year by the same court.
  • The ruling leaves open the possibility of an en banc rehearing or a Supreme Court petition.

What happened

In a 2‑1 decision issued on September 26, 2026, the DC Circuit concluded that the Trump administration did not exceed its authority under the Supply Chain Security Act. The court rejected Anthropic’s petition for review, stating that the Secretary of Defense acted within statutory limits. The judgment cited the “deeply sobering prospect of overly constrained AI models shutting down unexpectedly and thus causing important military operations to fail.”

Anthropic had sued the administration in March after the Pentagon ordered all federal agencies to cease using its products and prohibited defense contractors from purchasing them. The company argued that the blacklist violated 10 U.S.C. § 3252, which requires a malicious adversary motive for a supply‑chain risk designation. The appeals court, however, applied the broader 41 U.S.C. § 4713 framework, which does not demand proof of malicious intent. Anthropic’s spokesperson told CNBC, “We respectfully disagree with the court’s decision,” and indicated the firm may seek an en banc review or Supreme Court review.

The backstory

The dispute began when the Department of Defense, citing concerns that Anthropic’s models could hallucinate inappropriate targets, demanded that the company enable specific Claude capabilities for military applications. Anthropic refused, prompting the administration to invoke supply‑chain risk statutes. A Northern District of California judge earlier ruled that the blacklist was unlawful under § 3252 because Anthropic lacked a malicious motive. That decision focused on the narrow definition of “adversary” in the statute, prompting the DOD to pursue a parallel challenge under the more expansive § 4713, which grants the DC Circuit exclusive jurisdiction over such procurement actions.

Legal implications

The ruling clarifies that the Pentagon can rely on the broader § 4713 authority to block vendors without demonstrating malicious intent. This interpretation widens the government’s toolkit for managing technology risk, especially as AI models become integral to defense systems. The decision also highlights the influence of judicial appointments; both majority judges were appointed by former President Trump, which may affect future supply‑chain litigation.

Business impact

For AI startups, the verdict signals that refusal to comply with specific government requests can trigger blacklisting, even absent evidence of wrongdoing. Companies that supply or integrate AI into defense contracts must anticipate stricter compliance demands and be prepared to negotiate feature activation. Conversely, firms willing to align their models with defense specifications may gain a competitive edge in a market where the Pentagon is a major purchaser.

What builders and businesses should do

  • Review contractual clauses for government‑mandated feature activation and assess technical feasibility.
  • Conduct a risk assessment of supply‑chain designations under both § 3252 and § 4713.
  • Establish a rapid response team to address potential blacklisting requests.
  • Engage legal counsel early to navigate statutory nuances and appellate options.

The bottom line

The DC Circuit’s affirmation of the blacklist expands the DOD’s leverage over AI vendors. Watch for any en banc motions or Supreme Court filings, which could reshape the legal standards governing technology procurement.

#AI #Legal #Defense #Supply Chain #Anthropic

Related reading