What happened
Google, an Alphabet Inc. subsidiary, signed an agreement with the U.S. Pentagon to deploy its artificial intelligence models in classified operations, according to The Information, though Reuters reported it could not independently verify this. This agreement was part of contracts signed in 2025 worth up to $200 million with major AI companies, including Google, OpenAI, and xAI, which also supply AI for sensitive government use. The Pentagon aims for full flexibility in defence applications, including weapons systems, without company-imposed restrictions. The contract specifies Google will assist with security settings and filters, while stating the AI is not for mass surveillance or autonomous weapons without human oversight, yet grants Google no veto power over government operational decisions.
Why it matters
This agreement expands the pool of frontier AI models available for U.S. defence applications, increasing the Pentagon's access to advanced capabilities. For AI providers, it establishes a precedent for government contracts that prioritise operational flexibility over company-imposed usage restrictions, potentially shifting the balance of control in military AI deployments. Procurement teams must now account for these evolving terms, which contrast with previous company stances on military AI use. This follows earlier reports of the Pentagon pressing AI companies like Anthropic and OpenAI for unrestricted access to their tools on classified networks.
