Pentagon Stops Using Anthropic AI Tools After Blacklisting Company

The US Department of Defence is no longer using artificial intelligence tools developed by Anthropic, according to an official statement provided to the BBC on Monday. This development marks the culmination of months of tension that began when the department designated the company as a supply chain risk on national security grounds. Defence Secretary Pete Hegseth initially announced in February that the Pentagon would phase out its use of Anthropic by late August, though implementation appears to have taken slightly longer than initially scheduled.

Multiple sources familiar with the matter, including former defence officials and contractors, revealed that Anthropic's Claude model remained in active use for research, analysis, and intelligence gathering until very recently. The tool had even been utilised in military operations against Iran, serving as a critical asset for sorting through vast quantities of operational data. Anthropic declined to comment when asked about the Pentagon's confirmation that all use had finally been brought to an end.

The friction between the military and the AI developer arose when the Pentagon pressed Anthropic to remove its safety guardrails from Claude, demanding unfettered access to the technology for military applications. Anthropic refused these requests, raising serious concerns regarding potential mass surveillance and the deployment of autonomous weapons. Following this refusal, Hegseth designated the company a national security supply-chain risk—a legal classification typically reserved for foreign firms deemed threatening to the United States.

In response to the designation, Anthropic filed a lawsuit against the Trump administration, describing the government's move as unprecedented and unlawful. As this legal battle proceeds through the courts, the Department of Defence has pivoted toward other major technology partners. The Pentagon has signed contracts with companies such as Google, xAI, and OpenAI, with OpenAI tools seeing particularly wide adoption across various military branches in recent months.

Technical complexities help explain the delay in fully eradicating Anthropic's tools from military systems. Claude was deeply embedded within the Pentagon's primary intelligence-organising platform, the Maven Smart System, which is operated by Palantir. Lauren Kahn, a senior research analyst at Georgetown's Center for Security and Emerging Technology and a former US defence official, noted that deeply integrated software cannot simply be unplugged overnight. The prolonged process of excision highlights the practical challenges modern militaries face when integrating and subsequently removing complex commercial AI infrastructure.