Over 100 Google DeepMind Employees Protest Military AI Use in Rare Open Letter

When the email arrived in Jeff Dean’s inbox, it carried the signatures of more than 100 of his own employees—researchers, engineers, and scientists who had helped build some of the most powerful AI systems on Earth. The message was direct: they did not want their work used for military surveillance or autonomous weapons, and they were willing to say so publicly.

“We believe that AI systems should not be used for mass surveillance or fully autonomous weapons. These applications fall outside the boundaries of responsible AI development.” — Google DeepMind Employees

The Letter That Broke the Silence

The letter, sent to Dean and other senior leaders at Google DeepMind this week, represents one of the most significant internal protests in the company’s recent history. The employees are specifically opposing the use of Gemini AI systems for U.S. surveillance operations and certain military applications—drawing a direct parallel to the ethical stand that Anthropic took against the Pentagon just days earlier.

The timing could not be more consequential. The AI industry is in the midst of an ethical reckoning, with companies facing intense pressure from both governments and their own workforces over how powerful AI systems should be deployed. Anthropic’s confrontation with the Trump administration—resulting in a federal blacklist—has shown the risks of taking a principled stand. Google’s employees are now testing whether their employer will follow a similar path.

The specific concerns outlined in the letter focus on two applications the employees consider beyond ethical boundaries: mass domestic surveillance and fully autonomous weapons systems. These are the same red lines that Anthropic tried to enforce in its Pentagon contract negotiations—and the same safeguards that OpenAI reportedly secured in its own military deal announced Friday.

The Corporate Response Challenge

Google’s historical position on military AI has been complicated. In 2018, the company faced a major employee revolt over its involvement in Project Maven, a Pentagon program that used AI to analyze drone footage. More than 3,000 employees signed a petition demanding the company end the contract, and several resigned in protest. Google eventually chose not to renew the contract and published AI principles that prohibited weapons applications.

The competitive pressure has shifted dramatically since then. OpenAI’s Pentagon deal, announced the same day as the DeepMind letter, shows that major AI companies are now willing to work with the military under certain conditions. Microsoft and Amazon have long maintained defense contracts. The question for Google is whether it can afford to maintain its ethical stance while competitors secure lucrative government business.

The employee leverage is significant. DeepMind researchers are among the most sought-after talent in AI. If a meaningful portion were to leave over ethical concerns, it would represent a significant blow to Google’s AI ambitions. The letter is both a moral statement and a bargaining chip—employees using their collective influence to shape company policy.

“This is a pivotal moment for the AI industry. The decisions companies make now about military applications will define their relationship with both governments and their own workforces for years to come.” — AI Ethics Researcher

The Broader Industry Context

The DeepMind letter arrives at a moment of intense scrutiny for AI companies. Anthropic’s stand against the Pentagon resulted in President Trump calling the company “leftwing nut jobs” and ordering all federal agencies to cease using its technology. Defense Secretary Pete Hegseth designated Anthropic a supply chain risk to national security—a move that could have lasting implications for the company’s enterprise business.

Meanwhile, AI safety researchers are resigning from major labs at an unprecedented rate. At both OpenAI and Anthropic, prominent researchers have departed in recent weeks, warning that the race toward more powerful AI systems is outpacing safety research. The tension between commercial pressure and responsible development has never been more acute.

The political dimension is also escalating. New York State Assemblyman Alex Bores, who authored the nation’s first comprehensive AI safety law, now faces a $125 million super PAC funded by OpenAI cofounder Greg Brockman, Andreessen Horowitz, and Palantir’s Joe Lonsdale. The battle over AI governance is becoming a battle over political power.

For Google, the DeepMind letter presents a choice with no easy answers. Honor employee concerns and risk losing ground to competitors willing to work with the military? Or pursue defense contracts and face the possibility of an internal revolt that could damage both morale and recruiting? The decision will send a signal to the entire industry about where the lines are drawn.

The employees who signed the letter are waiting for a response. So is the rest of Silicon Valley.

This article was reported by the ArtificialDaily editorial team. For more information, visit The New York Times.