AI Giants Work Hand-in-Hand With the Pentagon, Contracts Reveal

The Pentagon and major artificial intelligence firms entered into secretive contracts worth up to $200 million in 2025, revealing a close partnership between the government and the tech industry. Documents obtained via a Freedom of Information Act lawsuit shed light on the extent of collaboration between the Department of Defense and Anthropic, Google, OpenAI, and xAI.

While the initial agreements focused on building AI tools to enhance military capabilities, the contracts also mandated extensive sharing of data, expertise, and intelligence. These agreements included provisions for AI labs to provide sensitive information, including benchmarks datasets and briefings on operational missions and threats, to the Pentagon.

In return, the companies were expected to offer feedback on military strategies, briefings on their technology, and training sessions for the military. The contracts also demanded tabletop exercises and risk forecasting to refine AI models for real-world challenges such as intelligence analysis, cybersecurity, and autonomous systems.

Despite the government's initial public statements emphasizing the importance of equipping warfighters with modern technology, the contracts reveal a much more intimate relationship between the military and AI firms.