Hundreds of artificial intelligence researchers employed at Alphabet Inc.'s Google have formally requested that Chief Executive Officer Sundar Pichai decline to permit the companys AI systems to be used for classified workloads supporting US defense missions. Organizers said the letter, received by Pichai on Monday, has secured more than 580 signatures.
According to the organizers, approximately two thirds of those who signed agreed to be identified by name, while roughly a third asked to remain anonymous. The letter raises concerns about ongoing negotiations between Google and the US Department of Defense and urges the company not to make its AI tools available for classified defense tasks.
In language provided by organizers, the letter says: "We are Google employees who are deeply concerned about ongoing negotiations between Google and the US Department of Defense." It continues: "As people working on AI, we know that these systems can centralize power and that they do make mistakes."
The timing of the employee appeal coincides with a separate legal confrontation between the Pentagon and Anthropic PBC over the use of AI for military purposes. In that dispute, the Pentagon is pursuing action to remove Anthropic and its Claude AI system from US defense supply chains, and it is said to be seeking alternative technology partners among major tech companies.
Google is not new to internal pushback over defense-related AI work. Employees raised objections in 2018 about the risks of AI being applied to warfare, a movement that led the company to limit some defense projects at that time. The current letter is the latest organized expression of concern from Googles AI workforce.
While the letters signatories press for a categorical refusal to run classified defense workloads, the broader implications for Googles commercial AI business and potential government partnerships remain tied to the separate Pentagon-Anthropic dispute and any subsequent procurement decisions. Organizers provided the letter to company leadership on Monday, and the group has sought to draw attention to the possible consequences of routing classified defense work through large AI systems.