Anthropic and the Pentagon Clash Over AI Control and Security Risks

The Department of Defense (DoD) and Anthropic are locked in a dispute over the control and potential security risks associated with the AI developer’s Claude model, escalating tensions over the use of artificial intelligence in national security. The conflict centers on Anthropic’s refusal to allow for mass domestic surveillance and the development of fully autonomous weapons systems using its technology.

Pentagon Labels Anthropic a Supply Chain Risk

In February 2026, Secretary of War Pete Hegseth directed the DoD to designate Anthropic as a supply chain risk , a move that could significantly restrict the company’s involvement in defense contracts. This action followed stalled negotiations regarding Anthropic’s requested exceptions concerning the use of Claude. The designation prevents the Department of Defense from using the company’s software, including through contractors, over the coming months, and other federal agencies are also abandoning Claude.

Anthropic’s Position: Safeguarding Democratic Values

Anthropic maintains that it has acted proactively to deploy its models to the Department of War and the intelligence community, being the first frontier AI company to do so in classified networks and at National Laboratories . However, the company has drawn a line regarding applications it believes could undermine democratic values. Specifically, Anthropic objects to the use of Claude for mass domestic surveillance and fully autonomous weapons, arguing that current AI models are not reliable enough for the latter and that the former violates fundamental rights .

Claims of Control and “Kill Switch” Concerns

The DoD has expressed concerns that Anthropic could disrupt military operations by disabling access to Claude or introducing harmful updates if it disagreed with certain uses of the technology. However, Thiyagu Ramasamy, Anthropic’s head of public sector, asserted that the company lacks the ability to unilaterally impact ongoing military operations. “Anthropic has never had the ability to cause Claude to stop working, alter its functionality, shut off access, or otherwise influence or imperil military operations,” Ramasamy wrote in a court filing . He further stated that Anthropic does not maintain a “back door” or remote “kill switch” and cannot access user prompts or data.

Legal Battles and Contract Proposals

Anthropic has filed two lawsuits challenging the constitutionality of the DoD’s ban and seeking an emergency order to reverse it . The company proposed a contract on March 4th guaranteeing it would not control or veto lawful Department of War operational decision-making . Anthropic also indicated a willingness to address concerns about autonomous weapons. Despite these offers, negotiations ultimately failed.

Mitigation Efforts and Cloud Provider Role

The Department of Defense, in court filings, stated it is taking steps to mitigate the supply chain risk posed by Anthropic by working with third-party cloud service providers, such as Amazon Web Services, to ensure Anthropic leadership cannot make unilateral changes to Claude systems currently in place . A hearing in one of the cases is scheduled for March 24 in federal district court in San Francisco, where a judge could decide on a temporary reversal of the ban.

Key Takeaways

• The DoD designated Anthropic a supply chain risk due to concerns over potential control and security vulnerabilities.
• Anthropic opposes the use of its AI for mass domestic surveillance and fully autonomous weapons.
• Anthropic asserts it lacks the ability to disrupt ongoing military operations through its AI models.
• The company is pursuing legal action to challenge the DoD’s ban.
• The DoD is working with cloud providers to mitigate risks associated with Anthropic’s technology.