OpenAI Reaches Pentagon Agreement with Technical Safeguards

OpenAI CEO Sam Altman announced that OpenAI has reached an agreement with the U.S. Department of Defense allowing the department to use its AI models in a classified network. The contract includes technical safeguards and principles addressing concerns around surveillance and autonomous weapon systems.

Altman stated that the agreement incorporates prohibitions on domestic mass surveillance and ensures human accountability for the use of force. He said these principles are reflected both in law and policy and have been formally included in the contract. OpenAI will also build technical safeguards to monitor model behavior and will deploy engineers to work alongside defense personnel to maintain system safety.

The announcement follows a failed negotiation between the Pentagon and Anthropic, which declined to permit unrestricted use of its models for military purposes. After that breakdown, federal officials criticized Anthropic’s stance, and the company was labeled a potential supply-chain risk.

Altman added that OpenAI asked the Department of Defense to offer the same terms to all AI companies and expressed a preference for resolving disagreements through mutual agreements rather than legal or regulatory measures.