OpenAI has announced a formal agreement with the U.S. Department of War to deploy its artificial intelligence models on classified military networks, stressing firm safety commitments and operational guardrails.
In a blog post titled “Our agreement with the Department of War,” the San Francisco-based startup explicitly prohibits the use of its AI for domestic surveillance of U.S. persons or by intelligence agencies such as the National Security Agency without a separate contract.
It also sets clear red-line limits against using the technology to independently direct autonomous weapon systems or make high-stakes automated decisions.

“In our agreement, we protect our red lines through a more expansive, multi-layered approach. We retain full discretion over our safety stack, we deploy via cloud, cleared OpenAI personnel are in the loop, and we have strong contractual protections. This is all in addition to the strong existing protections in U.S. law,” OpenAI said.

The move follows weeks of industry tension, including halted negotiations between the Pentagon and rival AI lab Anthropic.
OpenAI said it hopes the agreement can serve as a model for future collaboration between government and AI labs, and urged similar terms be offered across the sector.