In a move that has sparked internal opposition, Google has entered into an agreement to supply the Pentagon with its artificial intelligence technology for classified purposes, as reported by The Information. This deal permits the U.S. Department of Defense to utilize Google’s AI models for any lawful governmental objective, according to sources speaking with The New York Times. The contractual language is similar to those recently established between the Pentagon and OpenAI and xAI.
A spokesperson for Google told The New York Times, “We are proud to be part of a consortium of leading A.I. labs and technology firms providing AI services and infrastructure in support of national security.” They reiterated their commitment to preventing the use of AI in domestic mass surveillance or autonomous weaponry without proper human oversight.
Google did not respond to comments solicited by Decrypt regarding the deal’s specifics. Prior to the agreement, hundreds of Google employees penned an open letter to CEO Sundar Pichai, urging against making its AI systems available for Pentagon work. The letter expressed concerns that “AI should benefit humanity and not be used in harmful ways,” advocating a rejection of classified projects to prevent unknowingly supporting unethical applications like lethal autonomous weapons or mass surveillance.
The employees highlighted the potential for AI errors and power centralization, emphasizing Google’s duty to mitigate its most dangerous uses. They cautioned that an erroneous decision could severely tarnish Google’s reputation and global standing.
Since January, when Defense Secretary Pete Hegseth called for integrating technology across the military, the Pentagon has been actively pursuing agreements with major AI companies. This effort underscores a growing rift between military objectives and some AI developers over warfare applications of the technology.
In March, after CEO Dario Amodei refused unrestricted access to its AI models, the Pentagon identified Anthropic as a “supply chain risk,” restricting its federal collaboration until it sued for reconsideration.
Despite employee concerns, Google is advancing with the Pentagon contract, aligning with the Defense Department’s broader strategy of incorporating AI into classified operations. Hegseth emphasized the necessity of technological supremacy in a speech at Elon Musk’s Starbase in January, stating that the U.S. must quickly integrate leading AI models across all networks within the department.