OpenAI has released additional details about its agreement with the U.S. Department of Defense after criticism that the deal was rushed and raised concerns about the military use of artificial intelligence.

The disclosure followed the collapse of negotiations between another AI company and the Pentagon, which led the U.S. government to suspend use of that company’s technology.

OpenAI subsequently announced its own agreement to deploy AI models in classified environments.

In a public statement, OpenAI said its models cannot be used for mass domestic surveillance, autonomous weapons systems, or high-stakes automated decision-making. The company emphasized that these restrictions are enforced through technical design, deployment architecture, and contractual controls.

OpenAI added that its models will be deployed via cloud-based systems with strict oversight, preventing direct integration into weapons or surveillance hardware. The company argued that deployment structure is more effective than policy language alone.

Despite the assurances, critics have questioned whether existing legal frameworks could still allow broad data collection.

OpenAI maintains that multiple layers of safeguards significantly reduce such risks.

Source: TechCrunch