The US Department of Defence has officially stopped using Anthropic’s AI tools after designating the company a national security risk. The move comes months after the Pentagon issued its warning, with Secretary of Defence Pete Hegseth setting a deadline for late August.

Despite the blacklisting, Anthropic's Claude model was reportedly still in use within the department for research, analysis, and intelligence gathering, including in military operations against Iran. The firm faced legal battles over the designation, with Anthropic suing the Trump administration.

Anthropic's tools were deeply integrated into the Pentagon's Maven Smart System, a primary platform for organising intelligence data. Removing the AI proved challenging, with sources noting the difficulty in quickly removing such integrated technologies from the system.

The Pentagon's shift towards other major tech companies, including Google and OpenAI, highlights the ongoing competition in the AI space. As relations with the White House appear to warm, Anthropic's boss Dario Amodei met with President Trump, though concerns over AI risks remain.