Pentagon Ceases Utilization of Anthropic AI Services Over Security Concerns

Original summary · What happened

The United States Department of Defense has terminated all operational usage of artificial intelligence applications provided by Anthropic, according to a department official. The move formalizes an earlier national security determination made in February by Defense Secretary Pete Hegseth, which classified the artificial intelligence developer as a supply chain risk. While the Pentagon had initially targeted late August to phase out the vendor, the agency confirmed that deployment of the company's tools has now officially stopped across all departmental activities.

Anthropic was the initial frontier artificial intelligence firm permitted to integrate its models into classified United States military environments, maintaining government deployments dating back to 2024. Sources familiar with the operational history indicated that the developer's Claude model remained actively deployed as recently as last week, serving personnel in defense research, data sorting, intelligence processing, and direct military operations against Iran. Anthropic spokespersons declined to provide commentary regarding the Pentagon's official termination announcement.