OpenAI has signed a new agreement with the United States Department of Defense to deploy its artificial intelligence technology on classified military networks, just hours after the Trump administration ordered federal agencies to stop using rival company Anthropic’s tools.
The move followed a dramatic escalation in a dispute between the Pentagon and Anthropic over how AI systems could be used by the military. The administration labeled Anthropic a national security “supply-chain risk” after the company refused to allow unrestricted use of its AI, including potential applications in mass surveillance and autonomous weapons systems.
Soon after the ban, OpenAI CEO Sam Altman announced that his company had reached a deal with the Pentagon under conditions that include safety protections. According to Altman, the agreement prohibits domestic mass surveillance and requires human responsibility in the use of force, including decisions involving weapons systems.
Anthropic had sought similar assurances but insisted they be explicitly written into its contract, leading to a breakdown in negotiations. U.S. officials argued they required broader access for lawful military operations, while the company maintained ethical concerns about how advanced AI could be deployed.
The development highlights a growing divide within the technology sector over military uses of artificial intelligence. Some industry leaders and safety advocates supported Anthropic’s stance, warning about the risks of weaponized AI, while others backed the government’s national security priorities.
OpenAI’s agreement now positions the company as a major AI partner for U.S. defense projects, potentially reshaping competition in the rapidly expanding military-AI market. Analysts say the episode underscores intensifying global competition to control advanced AI systems and the increasing role of private tech firms in national security.
Anthropic has indicated it may challenge the government’s decision in court, meaning the broader legal and ethical battle over AI warfare and surveillance is far from over.
