The Pentagon is no longer using Anthropic's artificial intelligence tools, a U.S. Department of Defence official told the BBC — drawing a line under months of tension between the military and the AI company months after it was designated a national-security “supply chain risk.”
Defence Secretary Pete Hegseth labelled Anthropic a supply-chain risk in February 2026 after the company refused to remove safety guardrails from its Claude model — restrictions that bar the use of its technology for fully autonomous lethal weapons and mass domestic surveillance. The March designation cancelled Anthropic's military contracts and barred defence contractors from using its models.
The Pentagon had said it would stop using Claude by late August, but multiple people told the BBC the company's tools were still in use as recently as last week — for research, analysis, intelligence gathering and military operations against Iran. Anthropic has supplied the Pentagon since 2024 and was the first advanced AI company deployed in classified government work.
Anthropic called the blacklisting unlawful and sued; a divided D.C. Circuit panel upheld the designation 2-1 on September 25. The company says the fallout has cost it billions in lost business as it heads toward a potential public offering — and the military has shifted its AI work further toward vendors including Google, xAI and OpenAI.
