Google Signs Classified AI Deal with Pentagon, Overruling 600 Employee Protests

Google has signed a classified AI deal with the Pentagon, granting the U.S. Department of Defense access to its Gemini AI systems on classified networks for "any lawful government purpose." The deal was finalized on April 28, 2026.

The agreement centers on API access, enabling the Pentagon to integrate Google's AI directly into its software infrastructure without requiring custom model development. A clause in the contract specifies the AI "is not intended for, and should not be used for, domestic mass surveillance or autonomous weapons (including target selection) without appropriate human oversight and control." However, the deal does not give Google authority to veto lawful government decisions.

The move follows Anthropic's refusal to sign a similar deal without stronger safeguards — placing Anthropic in a separate legal dispute. OpenAI had already granted the Pentagon access to its models. Google's deal now positions it alongside OpenAI as a primary AI supplier to the U.S. military.

Inside Google, the decision triggered significant backlash. More than 600 employees sent a letter to CEO Sundar Pichai urging him to reject the classified agreement, arguing it was the only way to prevent Google's AI from being weaponized. In a separate development, Google reportedly exited a $100 million drone swarm program over ethical concerns — even as it signed the classified deal.

The arrangement underscores the growing strategic importance of AI to national defense, and the deepening tensions between tech companies' commercial ambitions and their employees' ethical limits. Full details at TechCrunch.