Google Negotiates Classified Gemini Deal With Pentagon

Sources say the tech giant is in talks to deploy its AI models in classified military networks, raising questions about surveillance and autonomous weapons limits.

Apr. 16, 2026 at 10:07pm

Got story updates? Submit your updates here. ›

A highly detailed, glowing 3D macro illustration of a complex circuit board with neon cyan and magenta lights pulsing through the interconnected components, conceptually representing the integration of advanced AI technology into classified military infrastructure.As Google and the Pentagon negotiate the terms for deploying the tech giant's AI models in classified military networks, the integration of commercial algorithms into secure government systems raises complex questions about surveillance, autonomous weapons, and the future of AI governance.Washington Today

Google is in negotiations with the Pentagon to allow the Defense Department to run the company's Gemini AI models inside classified systems, according to people familiar with the talks. The discussions center on contract language that would limit how the models can be used, including prohibiting their use for domestic mass surveillance and fully autonomous lethal weapons. However, the Pentagon is pushing for broad 'all lawful uses' wording that would give it more operational flexibility, putting the company's public AI safety promises under direct national security pressure.

Why it matters

This high-stakes tech courtship highlights the growing tension between commercial AI vendors' public commitments to ethical AI and the Pentagon's demand for operational flexibility in classified military networks. The outcome of these negotiations could set a powerful template for how commercial AI models operate inside secure government systems for years to come.

The details

Google has proposed contract language that would prohibit its Gemini AI models from being used for domestic mass surveillance and from controlling autonomous weapons without 'appropriate human control.' Company negotiators are also weighing an arrangement that would let the Pentagon deploy Gemini models inside classified environments, extending their reach far beyond current unclassified uses. Defense officials, however, have pushed for broad 'all lawful uses' wording that they argue is necessary to move quickly during crises. This tug-of-war over usage guardrails could reshape how the federal government writes AI contracts going forward.

  • In March 2026, the Pentagon launched GenAI.mil and made Google's Gemini the first model available on its unclassified networks.
  • The Pentagon's use of the rare "supply-chain risk" label against Anthropic has already triggered litigation and rattled procurement norms.

The players

Google

A multinational technology company that develops and provides internet-related services and products, including the Gemini AI model that is at the center of these negotiations with the Pentagon.

The Pentagon

The headquarters of the United States Department of Defense, which is in negotiations with Google to deploy the Gemini AI model within classified military networks.

Anthropic

An AI research company that previously faced a public standoff with the Pentagon over usage guardrails for its technology, leading to the Pentagon labeling it a "supply-chain risk" - a rare move that has triggered litigation and rattled procurement norms.

Got photos? Submit your photos here. ›

What they’re saying

“We're starting with unclassified [networks] and will move to classified [systems] next.”

— Emil Michael

What’s next

Legal experts say the outcome of these negotiations between Google and the Pentagon could reshape how the federal government writes AI contracts for years to come. The Pentagon's use of the rare "supply-chain risk" label against Anthropic has already triggered litigation and rattled procurement norms, setting the stage for further changes to come.

The takeaway

This case highlights the growing tension between commercial AI vendors' public commitments to ethical AI and the Pentagon's demand for operational flexibility in classified military networks. The final agreement, if any, between Google and the Pentagon could set a powerful template for how commercial AI models are allowed to operate inside secure government systems, with significant implications for the future of AI governance and national security.