The US military is urging Anthropic to rapidly revise its AI moral guidelines, sparking a high-stakes dispute over the ethics of artificial intelligence. With a Friday deadline looming, the Pentagon has threatened to designate Anthropic as a ‘supply chain risk’ or invoke the Defense Production Act if it refuses to comply.
The Pentagon and Anthropic are locked in a high-stakes dispute over the company’s ethical guidelines for artificial intelligence, with Defense Secretary Pete Hegseth issuing a Friday deadline for Anthropic to abandon its restrictions or face severe consequences. ‘or else’
Hegseth met with Anthropic CEO Dario Amodei on February 23, 2026, to present two options: either the company must remove its ethical safeguards or the Pentagon will designate Anthropic as a ‘supply chain risk‘ or invoke the Defense Production Act to enforce compliance. The latter designation, typically reserved for foreign adversaries, would bar Anthropic from U.S. government contracts and require other Pentagon contractors to sever ties with the company, according to Politico.
Anthropic has maintained two core prohibitions it refuses to compromise on: it will not allow its AI model Claude to be used for mass domestic surveillance of U.S. citizens or for autonomous weapons systems where AI makes targeting decisions without human oversight. ‘The company has no intention of easing these restrictions despite the Pentagon’s pressure’
Reuters reported that the company has no intention of easing these restrictions despite the Pentagon‘s pressure.
The dispute underscores a broader tension between U.S. military interests and corporate ethical frameworks in the development of advanced AI technologies.
The standoff carries particular weight because Anthropic is currently the only frontier AI laboratory with classified Department of Defense access, leaving the Pentagon without an immediate alternative. While the DOD has reportedly reached a deal to use xAI’s Grok in classified systems, this arrangement is not yet operational, creating significant leverage for Anthropic in the standoff.
The Pentagon‘s aggressive posture appears driven by this lack of redundancy and its failure to comply with a -era National Security Memorandum directing federal agencies to avoid dependence on a single classified-ready AI system.
The meeting between Hegseth and Amodei follows months of escalating tensions. ‘s refusal to compromise on its ethical red lines has prompted the Pentagon to consider measures that would limit the company’s influence over U.S. defense systems. The Defense Production Act, a tool used to prioritize national security needs, could force Anthropic to align its practices with Pentagon directives or face exclusion from critical contracts.
This scenario reflects a broader conflict between innovation and regulation in the AI sector, with national security agencies seeking to balance technological advancement with ethical accountability.
Industry analysts note that the dispute highlights the challenges of integrating AI into military operations while adhering to ethical constraints. Anthropic’s position reflects a growing trend among tech companies to prioritize privacy and safety over unchecked military applications, a stance that has drawn both support and criticism from policymakers.
As the Pentagon prepares to enforce its demands, the outcome could set a precedent for how government agencies engage with private-sector AI developers in the future.
