Pentagon Blacklists Anthropic as Supply-Chain Risk: Legal Battle Over AI Use in Defense

After months of talks over Anthropic’s Claude models for weapons and surveillance, Pentagon labeled startup a “supply-chain risk” in March 2026. Secretary Pete Hegseth framed it as a national-security precaution, prompting Donald Trump to halt agency use of Anthropic’s technology. DoD, with DOJ attorney Sharon Swingle, faces Anthropic, led by CEO Dario Amodei and counsel Kelly Dunbar, before D.C. Circuit panel with Judges Karen Henderson, Neomi Rao and Gregory Katsas.

CNBC

In March 2026 Anthropic sued Pentagon and Secretary Hegseth in Washington, San Francisco over the supply-chain risk label. The D.C. Circuit denied a request in April to block the designation pending appeal. On May 19 2026 the three-judge panel—Henderson, Rao, Katsas—heard arguments, examining jurisdiction, procedural compliance and less intrusive alternatives before issuing an opinion.

The Independent

Anthropic posted $30 billion in annualized revenue, up from about $10 billion a year earlier, and is pursuing a funding round that could value the company at $900 billion, surpassing OpenAI’s $380 billion valuation in February 2026.

CNBC

DoD, DOJ attorney Sharon Swingle, said Anthropic’s models could be “encoded with limitations,” an “untenable national-security risk” and warned that failure would cause “catastrophic national-security consequences.” Anthropic counsel Kelly Dunbar countered that supply-chain risk designation “defied congressionally mandated procedures, exceeded statutory limits and violated the Constitution,” and no “record evidence” of restrictive code.

The Independent

Judge Karen Henderson called Pentagon’s action “a spectacular overreach” and said she saw no evidence of a supply-chain risk, noting the department had not shown a “significant supply-chain risk.” A San Francisco judge described the designation an “Orwellian notion” used to leverage a dispute.

The Independent

DOJ asserts Anthropic can “interfere with and even prevent” Pentagon use of its AI model in operations and warns of “catastrophic” failure consequences. Anthropic counters there is no “record evidence” of such capability, leaving a factual gap for court.

CNBC

• “To me, this is just a spectacular overreach by the (Defense) Department,” — Judge Karen Henderson, D.C. Circuit

The Independent

• “For the first time ever, the secretary turned a powerful national security authority against an American company, and he did so to gain leverage in a contract dispute,” — Kelly Dunbar, Anthropic attorney

The Independent

• “Anthropic CEO Dario Amodei said the company had "no choice" but to challenge the supply chain risk designation in court.” — Dario Amodei, CEO, Anthropic

CNBC

• “It's undisputed that the failure of the model in active military operations could have catastrophic national-security consequences and put service members' lives at risk,” — Sharon Swingle, DOJ attorney

The Independent

The panel will issue an opinion in the coming weeks. The San Francisco injunction keeps Anthropic’s models available to non-DoD agencies, and the case may shape future DoD AI procurement rules and use of supply-chain risk designations against domestic firms.

CNBC