Full Breakdown
Pentagon Upheld Blacklisting of AI Firm Anthropic
By Drooid · · How we work
Core Ruling and Immediate Impact
On September 25, 2026, a three-judge panel of the U.S. Court of Appeals for the District of Columbia Circuit affirmed the Department of War’s designation of Anthropic PBC as a “supply-chain risk.” The 2-1 decision bars the Pentagon and its contractors from using Anthropic’s Claude models and prevents new defense contracts with the company. The ruling resolves the D.C. Circuit portion of Anthropic’s challenge to the March 2026 designation but leaves a parallel California decision in place.
Background & Context
The Pentagon invoked the Federal Acquisition Supply Chain Security Act of 2018 (FASCSA), a law originally crafted to exclude firms with ties to hostile foreign actors. Anthropic, a San Francisco AI startup, had refused to remove contractual safeguards that prohibit Claude from being employed in fully autonomous lethal weapons or mass domestic surveillance. The Department argued those safeguards could cause Claude to fail to perform lawful military tasks, constituting a national-security risk under the statute’s broad language.
Timeline
- May 19, 2026 – D.C. Circuit panel heard oral arguments.
- March 3, 2026 – Secretary of War Pete Hegseth issued a written determination that Claude posed a significant supply-chain risk.
- March 4, 2026 – Anthropic received formal notice of the designation.
- March 6, 2026 – Department CIO ordered removal of Anthropic products from all DoD systems within 180 days.
- April 8, 2026 – Court denied a stay of the designation.
- June 24, 2026 – Cases consolidated after a second petition.
- August 27, 2026 – California federal judge struck down a separate Pentagon designation under a different statute.
- September 25, 2026 – D.C. Circuit upheld the FASCSA designation.
Data & Statistics
- Anthropic estimates the blacklisting has cost the company billions of dollars in lost business ahead of a planned IPO.
- The company’s valuation has been reported at nearly $1 trillion, with IPO fundraising targets of up to $100 billion.
- The Pentagon’s 2025 contract with Anthropic carried a $200 million ceiling, though no payments were made before the ban.
Conflicting Reports & Gaps
Two federal courts have issued opposing rulings on separate designations of Anthropic. The D.C. Circuit upheld the 2018 FASCSA designation, while the Northern District of California found a parallel designation under a different law unlawful. The legal community lacks a definitive interpretation of whether contractual safety guardrails qualify as “malicious interference” under the statute, leaving the issue unresolved pending possible en banc or Supreme Court review.
Verbatim Quotes
- “The Department had ample support for its conclusion that the continued integration of Claude into the Department's information systems, by the Department or its contractors, presented a statutorily covered national-security risk,” — Judge Gregory Katsas
- “The hammer of justice has smashed [Anthropic] arguments,” — Emil Michael, Pentagon CTO
What’s Next
Anthropic’s legal team plans to seek an en banc rehearing of the D.C. Circuit decision and, if denied, to petition the Supreme Court. The Pentagon has indicated no intention to alter the designation pending further judicial review.
