Google signs classified AI deal with Pentagon, allowing use for any lawful government purpose

Alphabet's Google joined OpenAI and xAI in signing an agreement with the U.S. Department of Defense to supply AI models for classified work, according to The Information. The Pentagon signed agreements worth up to $200 million each with major AI labs in 2025, including Anthropic, OpenAI, and Google.

Google's agreement requires the company to help adjust its AI safety settings and filters at the government's request. The contract specifies that the AI system should not be used for domestic mass surveillance or autonomous weapons without appropriate human oversight, though the agreement does not confer any right for the government to veto lawful operational decisions.

The deal is an amendment to Google's existing contract with the U.S. government, handled through Google Public Sector, the unit managing U.S. government business.