Google grants U.S. military access to AI for classified networks

Google has officially granted the U.S. Department of Defense access to its artificial intelligence for classified networks, which are secure government communication systems. The agreement essentially allows all lawful uses of the technology on these secure networks. By entering into this partnership, Google joins competitors OpenAI and xAI, both of which previously signed deals to provide the U.S. military with access to their AI models on classified networks.

This agreement follows a public standoff between the U.S. Department of Defense and Anthropic. Anthropic refused to grant the military the same terms because the Pentagon sought unrestricted use of the AI, whereas Anthropic wanted guardrails to prevent its technology from being used for domestic mass surveillance and autonomous weapons. In response to this refusal, the Department of Defense designated Anthropic a “supply-chain risk”—a designation used by the department for entities deemed a security threat. Anthropic subsequently challenged this designation in court, and last month, a judge granted Anthropic an injunction against the supply-chain risk designation while the lawsuit proceeds.

Google’s contract reportedly contains language stating that the company does not intend for its AI to be used for domestic mass surveillance or in autonomous weapons, which is similar to contract language previously signed by OpenAI. However, The Wall Street Journal reported that it is unclear whether these contract provisions regarding surveillance and weapons are legally binding or enforceable, as the agreement essentially permits all lawful uses of the technology.

The decision to proceed with the deal has caused internal friction at Google. A group of 950 Google employees signed an open letter asking the company to follow Anthropic’s lead and refuse to sell its AI to the Department of Defense without guardrails. Google did not respond to a request for comment regarding the contract or the employee opposition.

Why it matters

The move highlights the growing divide between AI labs willing to work with the U.S. military and those prioritizing strict guardrails, even as internal employee pressure mounts.