Over 600 Google DeepMind Employees Sign Letter Demanding Ban on Google AI Use for Classified Military Operations

More than 600 employees of Google DeepMind and Google Cloud have sent a letter to CEO Sundar Pichai, demanding the company avoid any agreement that would allow the Department of Defense (DoD) to use Google’s AI models in classified military operations. The letter, reported on April 27, 2026, follows earlier reports that Google was negotiating a contract to deploy its Gemini AI model in a classified environment for the Pentagon. The DoD reportedly insisted on allowing ‘all legitimate uses,’ while Google proposed clauses to prohibit large-scale domestic surveillance and fully autonomous weapons without human oversight.

The employees argue that using Google’s AI for classified military work would violate the company’s stated AI principles and risk enabling lethal autonomous weapons and mass surveillance. They wrote that as AI developers, they have a responsibility to expose and prevent ‘the most unethical and dangerous uses’ of the technology. This internal revolt echoes a similar standoff between Anthropic and the DoD, where the Pentagon threatened to cut ties after Anthropic’s CEO refused to allow unrestricted military use. As of press time, Google has not commented on the letter.