The United States Department of Defense has officially ended its use of Anthropic’s artificial intelligence tools, an official confirmed to the BBC. This action comes months after the company was designated a national security supply chain risk, effectively banning its technology from government operations.
Defense Secretary Pete Hegseth first labeled Anthropic a supply chain threat in February, setting a deadline for the Pentagon to cease using its products by late August. However, multiple sources indicated that Anthropic’s Claude model remained in active use as recently as last week for research, intelligence gathering, and military operations against Iran.
A Pentagon spokesperson stated, “The Pentagon has ceased the use of Anthropic products,” marking the first official confirmation of the full termination of services. The department had relied on Anthropic since 2024, making it the first advanced AI firm to deploy its tools in classified government work.
The conflict began when the Pentagon requested that Anthropic remove safety guardrails from its Claude model to allow unrestricted military access. Anthropic refused, citing ethical concerns regarding mass surveillance and autonomous weapons. In response, Hegseth invoked a legal provision typically reserved for entities from adversarial nations, branding the company a national security risk. Anthropic has described the designation as “unprecedented and unlawful” and filed a lawsuit against the Trump administration to overturn it.
Despite the political and legal standoff, integration issues delayed the transition. Sources revealed that Claude was deeply embedded within Maven, Palantir’s primary data platform used across the defense department for organizing intelligence and satellite imagery. Lauren Kahn, a senior analyst at Georgetown’s Center for Security and Emerging Technology, noted that removing such deeply integrated systems is rarely instantaneous.
“The fact that it took the Pentagon until now to remove all use of Claude from across the system is indicative that these things are not just plug and play,” Kahn said. “Once they become integrated it can be painful to remove them.”
As the legal battle proceeds, the Pentagon has shifted contracts to other major technology firms, including Google, xAI, and OpenAI. Reports suggest OpenAI tools have seen increased adoption in certain military branches. Meanwhile, relations between Anthropic and the White House appear to be improving, with President Trump recently meeting with Anthropic CEO Dario Amodei twice. Following their encounter, Trump praised Amodei, while the CEO expressed cautious optimism about collaborating on AI safety risks.
Leave a Reply