AI Ethics Battle: Anthropic Refuses Military AI Changes as OpenAI Signs Pentagon Deal

A dispute between artificial-intelligence company Anthropic and the United States Department of Defense has triggered a wider debate over military uses of AI, corporate ethics and the strategic positioning of leading technology firms.

The conflict began after Anthropic refused to remove safeguards in its AI system that limit how the technology can be used by the military. The company’s flagship model, Claude, includes policies prohibiting its use for domestic mass surveillance or fully autonomous weapons systems.

Pentagon officials had sought broader access to the system under a contract reportedly worth about $200 million, allowing what they described as “any lawful use” of the technology. Anthropic declined to lift those restrictions, arguing that the safeguards were essential to prevent misuse of advanced AI tools.

Government response

Following the breakdown in negotiations, Defense Secretary Pete Hegseth designated Anthropic a “supply-chain risk,” a classification that effectively removed the company from Department of Defense contracts and restricted the use of its technology in military programs.

The designation is unusual because the label has historically been applied to foreign companies considered national-security threats, rather than to U.S. technology firms.

President Donald Trump subsequently ordered federal agencies to stop using Anthropic’s AI systems after the contract dispute.

Anthropic has said it plans to challenge the decision in court, arguing the designation is unjustified and could set a precedent affecting the relationship between technology companies and government agencies.

OpenAI steps in

The same week the dispute became public, rival AI developer OpenAI announced a separate agreement with the Pentagon to deploy its models within classified government networks.

OpenAI chief executive Sam Altman said the agreement includes safety provisions intended to prevent domestic surveillance of U.S. citizens and maintain human oversight over the use of force in military systems.

The timing of the deal has intensified competition among leading AI developers as governments increasingly seek access to advanced machine-learning systems for intelligence analysis, logistics and battlefield decision-support.

Broader industry tensions

The episode highlights growing tensions between Silicon Valley companies and defense agencies over how artificial intelligence should be deployed in national-security operations.

Anthropic co-founder Dario Amodei has argued that AI systems should not be used to conduct mass surveillance or to operate weapons without meaningful human control.

Pentagon officials, however, have said that restrictions on military applications could hinder the United States in technological competition with rivals such as China, particularly in areas like autonomous defense systems and missile interception.

Despite the government designation, Anthropic said the impact on its broader business is limited, noting that the restriction primarily applies to the use of its models directly within Department of Defense contracts rather than across the wider commercial market.

Strategic implications

Industry analysts say the dispute underscores the strategic importance of AI companies in national security as governments seek access to increasingly powerful models. At the same time, it highlights the reputational and ethical risks companies face when their technologies are used in military or surveillance contexts.

The outcome of Anthropic’s legal challenge could shape future rules governing how private AI developers collaborate with governments and how much control companies retain over the uses of their technology.

For now, the confrontation illustrates a deeper divide in the rapidly evolving AI sector: whether the world’s most powerful algorithms should be tightly constrained by corporate safeguards or integrated more fully into military systems.