OpenAI strikes a deal with the Defense Department to deploy its AI models

Source: Engadget

Background

OpenAI has reached an agreement with the U.S. Defense Department—referred to by the company as the Department of War (DoW)—to deploy its AI models within the agency’s classified network. In a post on X, OpenAI chief Sam Altman highlighted two of the company’s core safety principles: a prohibition on domestic mass surveillance and a requirement for human responsibility when using force, including in autonomous weapon systems. Altman said these principles were incorporated into the contract with the DoW.

Agreement Details

• The DoW closed the deal shortly after former President Donald Trump ordered all government agencies to stop using Anthropic’s Claude and other Anthropic services.
• Defense Secretary Pete Hegseth had previously threatened to label Anthropic as a “supply chain risk” unless the company removed guardrails that prevent its technology from being used for mass surveillance or fully autonomous weapons.
• Altman stated that OpenAI is asking the government to extend the same safety terms to all AI companies it works with.
• Jeremy Lewin, Senior Official Under Secretary for Foreign Assistance, Humanitarian Affairs, and Religious Freedom, posted on X that the DoW contract “references certain existing legal authorities and includes certain mutually agreed upon safety mechanisms.”
• Both OpenAI and xAI, which previously signed a deal to deploy the Grok model in the DoW’s classified systems, agreed to these terms. Lewin described the arrangement as the same “compromise that Anthropic was offered, and rejected.”

Anthropic’s Response

Anthropic, which began collaborating with the U.S. government in 2024, refused to accept the DoW’s conditions. In a statement released hours before Altman’s announcement, Anthropic reiterated its stance:

“No amount of intimidation or punishment from the Department of War will change our position on mass domestic surveillance or fully autonomous weapons. We will challenge any supply chain risk designation in court.”

Technical Safeguards and Cloud Deployment

Altman added that OpenAI will implement technical safeguards to ensure its models behave as intended, aligning with the DoW’s safety requirements. The company will:

• Deploy engineers to work directly with the agency to verify model safety.
• Run the models on cloud networks only.

As noted by The New York Times, OpenAI is not yet hosted on Amazon Web Services (AWS), which the government currently uses. However, OpenAI has recently announced a partnership with Amazon to run its models on AWS for enterprise customers, potentially facilitating future deployment on the DoW’s preferred infrastructure.

Quote from Sam Altman

“Tonight, we reached an agreement with the Department of War to deploy our models in their classified network.
In all of our interactions, the DoW displayed a deep respect for safety and a desire to partner to achieve the best possible outcome.”
— Sam Altman (@sama) February 28, 2026