Google agrees with Pentagon to use its AI models for any lawful purposes

Google, according to The Information, has concluded a classified agreement with the Pentagon that permits the use of its AI models "for any lawful government purpose." The news came out just days after some employees called on Sundar Pichai not to allow military applications of these systems.

What the deal includes

This is not a public contract with a transparent list of services, but a closed agreement for the U.S. Department of Defense. The wording about "any lawful government purpose" appears maximally broad: it leaves the Pentagon room for a wide range of scenarios — from analytics and document processing to work with intelligence data, planning, and internal military operations. Due to its classified status, it remains unclear which specific Google models will be involved, what limitations are spelled out inside, and where the boundary of permissible use lies.

For Google, this is also a strategic turning point. The company has long tried to exercise caution on military AI projects following the loud scandal around Project Maven, but the market has changed. The largest AI laboratories are increasingly moving into the defense sector, because that's where the big budgets are, the long-term contracts, and direct access to tasks where artificial intelligence can become infrastructure rather than just an office tool. Essentially, Google is increasingly blurring the line between the civilian and government markets, and more often views its models as a platform for any major client.

Why employees are arguing

The news about the deal came out less than a day after Google employees called on the CEO to block the company's AI use by the Pentagon. Their main argument is the too-high risk that the models will be used not for routine automation, but in scenarios that could cause severe consequences for people. The internal conflict here is not only ethical but reputational: promises about "responsible AI" begin to look weaker when the details of the cooperation are hidden.

"for any lawful government purpose"

The problem is precisely in this formula. Legality in such contracts does not answer the question of how humane, controllable, or verifiable the system is. For critics this is not enough: they fear that it could cover tools for analyzing targets, processing intelligence data, selecting surveillance objects, and other sensitive operations, where a model error or weak operator oversight will have a real cost. That's why inside Google the argument is no longer about wording in press releases, but about the real boundaries of participation in military programs.

Who already works with the Pentagon

If the information is confirmed, Google will join other AI companies that have already concluded closed deals with the U.S. government. For the market this is an important signal: cooperation with defense is becoming not an exception but a new norm for major players. Not long ago such agreements looked like something isolated, but now a whole class of suppliers is forming, ready to adapt leading models to government tasks. This changes both the rules of competition and investor expectations.

• OpenAI has already concluded a separate agreement with the Pentagon on AI services.
• xAI also received a defense contract, which shows the government's willingness to work with more than one supplier.
• Anthropic was previously discussed as a possible partner, but according to earlier reports, was excluded after refusing to remove the internal restrictions that the Defense Department wanted removed.
• Competition is shifting from consumer chatbots to closed government implementations.
• For model developers this means not only revenue but also pressure on their own safety principles.

Against the backdrop of this race, Google risks a double effect. On the one hand, the company strengthens its position in the competition with OpenAI, Microsoft, and xAI for large government budgets. On the other — the broader the formulas in the contract and the fewer public explanations, the harder it will be for them to maintain the trust of employees, researchers, and users who expect clear red lines for military AI applications. The more aggressively Big Tech enters defense, the harder it becomes to maintain a neutral image.

What this means

The AI market is rapidly entering a phase where key customers are not only corporations but also government agencies. Google's deal with the Pentagon shows: the dispute over military use of models is no longer theoretical — now it affects both the business of the largest AI labs and what restrictions they are willing to really defend. For the industry this is a signal: the question is no longer whether government will use such systems, but who and on what terms will be able to control it.