OpenAI has announced its agreement with the Pentagon to deploy technology on the U.S. defense department's classified network. The deal includes enhanced safeguards to protect its use cases.
This development follows a directive from President Trump to cease work with AI firm Anthropic, which the Pentagon declared a supply-chain risk. Anthropic intends to challenge this designation legally.
OpenAI, a Microsoft-backed company, stated its agreement possesses more robust guardrails than previous classified AI deployments, including Anthropic's.
The contract with the Department of Defense enforces three critical restrictions: OpenAI technology cannot be used for mass domestic surveillance, to direct autonomous weapons systems, or for high-stakes automated decisions.
OpenAI outlined its multi-layered approach to enforce these "red lines," asserting control over its safety stack, deployment via cloud, involvement of cleared personnel, and strong contractual protections.
Past Pentagon agreements with major AI labs like Anthropic, OpenAI, and Google have been valued up to $200 million. The Defense Department aims to maintain flexibility in defense capabilities.
OpenAI cautioned that any breach of contract by the U.S. government could lead to termination, though they anticipate this will not occur. The company also stated that Anthropic should not be labeled a "supply-chain risk."