OpenAI Inks Classified Deal with DoD
OpenAI has reached a new agreement to deploy its AI models on secure, classified networks for the U.S. Department of Defense. The deal includes strict guardrails for data security and is designed for mission-critical use, signaling a deeper integration of commercial AI into national security.
The agreement is valued at up to $200 million and is part of a new initiative called "OpenAI For Government." The work, primarily taking place in the Washington, D.C. area, is expected to be completed by July 2026. The contract will focus on developing "prototype frontier AI capabilities" for both battlefield and internal administrative functions. This deal materialized shortly after the Pentagon dropped a similar contract with AI competitor Anthropic. Anthropic reportedly refused to remove safeguards that prevented its AI from being used for mass domestic surveillance and fully autonomous weapons, leading the DoD to seek a partner for "any lawful use." OpenAI's decision to take the contract sparked immediate backlash from the public and some of its own employees, who feared the technology could be misused. In response to the criticism, CEO Sam Altman admitted the rollout of the deal looked "opportunistic and sloppy" and announced amendments to the agreement. OpenAI has publicly stated its "red lines" for the partnership, asserting its technology will not be used for mass domestic surveillance, to direct autonomous weapons systems, or for high-stakes automated decisions. Following the public outcry, the contract was updated to explicitly prohibit its use for domestic surveillance and by intelligence agencies like the NSA without a separate agreement. Cleared OpenAI engineers will be directly involved in the government's use of the models to ensure compliance with these safety protocols. The company stated it retains full control over its safety features and that the deployment will be cloud-only, not on the "edge" in battlefield hardware, as an additional safeguard. The arrangement is managed by the Pentagon's Chief Digital & AI Officer (CDAO). Beyond the DoD, OpenAI's government initiative also consolidates existing work with other U.S. agencies, including NASA, the Air Force Research Laboratory, and the Treasury Department.
