June 17, 2025 — OpenAI has secured a major foothold in the U.S. defense sector, signing a $200 million contract with the Department of Defense to build advanced artificial intelligence tools aimed at improving cybersecurity, streamlining government operations, and strengthening national security infrastructure.
Announced this week by the Pentagon, the deal marks a significant shift in the government’s embrace of AI, tasking OpenAI with developing “prototype frontier AI capabilities” for both military and administrative applications. The contract is expected to run through July 2026 and will be executed primarily in the Washington, D.C. area.
According to the Department of Defense, the agreement includes AI-powered systems for data analysis, internal workflow automation, and proactive cyber defense—tools that may reshape how the military and related agencies process intelligence, manage operations, and respond to digital threats.
OpenAI confirmed the collaboration in a company blog post, describing it as the first in a new initiative designed to offer AI services across federal, state, and local governments. The firm emphasized that all deployments under this program will adhere to its internal safety protocols. Its current use policies prohibit deploying AI for weaponization, physical harm, or property destruction.
“This contract will bring OpenAI’s cutting-edge capabilities to vital areas like healthcare access for service members, data transparency in acquisition programs, and early detection of cyber threats,” the company said.
The move signals an evolution in OpenAI’s posture toward defense contracts. Just last year, the organization quietly dropped a longstanding policy that explicitly forbade military use of its models. In December 2024, it partnered with Anduril Industries to integrate AI into anti-drone defense systems—a clear sign the company was preparing for deeper involvement in national security.
OpenAI is not alone in this pivot. Its competitors have also been moving into the defense space. Anthropic recently unveiled a less restricted AI model tailored to intelligence work. Google revised its ethical commitments earlier this year, removing prior restrictions on harmful uses of AI. Meta has likewise opened its LLaMA model for use in government national security initiatives.
The new DoD partnership places OpenAI at the center of a growing race to integrate artificial intelligence into the core of U.S. defense strategy—raising important questions about safety, oversight, and the rapidly shifting boundaries between commercial AI and national security.