Pentagon Threatens Ending $200 Million Anthropic Deal Over AI Restrictions: Report

Last Updated:February 15, 2026, 18:11 IST

The US Department of Defense may end its 200 million dollar contract with Anthropic over disagreements on AI use for military purposes.

The US Department of Defense is considering ending its relationship with artificial intelligence company Anthropic after months of stalled negotiations over how the military can use the firm’s AI models, according to an Axios and Reuters report.

The Pentagon has been pushing four major AI developers, Anthropic, OpenAI, Google and xAI to allow the U.S. military to use their tools for “all lawful purposes," including weapons development, intelligence gathering and battlefield operations. Anthropic has not agreed to those broad terms, frustrating Pentagon officials after months of discussions.

Anthropic has maintained that its talks with US authorities have focused on usage policy issues, including clear limits on fully autonomous weapons and mass domestic surveillance safeguards, which it says are not linked to current military operations. The company has said it remains committed to supporting U.S. national security but insists on ethical guardrails around its AI deployment.

The dispute comes amid wider Pentagon efforts to push leading AI firms to make their tools available on classified defence networks with fewer restrictions, a move aimed at accelerating AI integration across military planning and operational systems. While OpenAI, Google and xAI have shown more flexibility on allowing broader use, Anthropic’s stance has emerged as a sticking point.

The tensions have been heightened by reports that Anthropic’s AI model Claude was used in a recent US military operation to capture Venezuelan leader Nicolás Maduro, though the company has declined to confirm specific operational details. That use facilitated through a partnership with data firm Palantir has underscored the ethical and policy challenges surrounding AI in defence contexts.

Also Read: US Used Anthropic’s Claude AI In Operation To Capture Venezuela’s Maduro: Report

Officials told Axios that ending or scaling back the Pentagon’s contract with Anthropic, reported to be worth around $200 million, is on the table. Any such move would likely require transitioning to alternative AI tools already being integrated into military networks.

The dispute reflects emerging tensions between rapid defence adoption of AI technologies and industry concerns about ethical safeguards, oversight and the potential misuse of advanced models in autonomous or surveillance capabilities.

News world Pentagon Threatens Ending $200 Million Anthropic Deal Over AI Restrictions: Report

Read More