Pentagon Stops Using Anthropic AI After Blacklisting Delay — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Pentagon officials confirmed to the BBC on Monday that the department "has ceased the use of Anthropic products," roughly two months after Secretary of Defence Pete Hegseth's late-August deadline and about eight months after he first blacklisted the company in February.
- Anthropic's Claude was still being used by the Pentagon as recently as last week for research, analysis, intelligence gathering, and military operations against Iran, according to multiple people familiar with the matter, including former US defence officials and contractors.
- Hegseth designated Anthropic a "national security supply-chain risk" — a label typically reserved for companies based in countries threatening the US — after Anthropic refused to strip safety guardrails from Claude, citing concerns about mass surveillance and autonomous weapons.
- Anthropic sued the Trump administration, calling the designation "unprecedented and unlawful," and a judge ruled on 28 August that the administration illegally retaliated against the company.
- Palantir's Maven Smart System is the Pentagon's primary platform for organising intelligence, and Claude was deeply embedded within it — analysts fed satellite imagery, drone footage, and other visual data into Claude to identify potential military targets and compile one-page briefs for the chain of command.
- OpenAI, Google, and xAI have all signed new Pentagon contracts during the Anthropic legal fight, with OpenAI tools more widely adopted in some military departments in recent months, according to two people familiar with defence department AI use.
- Trump met with Anthropic CEO Dario Amodei twice at the White House last week; Amodei said AI risk concerns remain "under discussion" but that "if we work with the president, we can win safely," signalling a possible thaw even as the lawsuit continues.
Why it matters: A federal judge has already ruled the Trump administration illegally retaliated against Anthropic, giving the company legal leverage while competitors OpenAI, Google, and xAI absorb the Pentagon contracts Anthropic is losing. The months-long delay in fully removing Claude — which sources say continued running in operations against Iran after the August deadline — exposes how hard it is to swap out AI tools once they're wired into Palantir's Maven system, raising the cost of any future government attempt to blacklist a single AI vendor.
Ask SkimNews
