600 Google staff urge CEO to reject classified US military AI contract

Over 600 Google staff urge CEO to reject classified US military AI contract Sign up now: Get ST's newsletters delivered to your inboxGoogle is in negotiation with the US Department of Defence to deploy its flagship Gemini AI model in classified settings, which is raising concerns of individuals being profiled.PHOTO: REUTERSPublished Apr 28, 2026, 09:46 AMUpdated Apr 28, 2026, 11:00 AMListenWASHINGTON – More than 600 Google employees demanded on April 27 that the company reject a proposed Pentagon deal that would allow its artificial intelligence technology to be deployed in classified military operations, a statement said.The letter, addressed to chief executive Sundar Pichai and signed by workers from Google DeepMind, Cloud and other divisions, comes as the tech giant is in active negotiations with the US Department of Defence to deploy its flagship Gemini AI model in classified settings.More than 20 directors, senior directors and vice-presidents were among the signatories, the statement added.“Classified workloads are by definition opaque,” one organising employee, who was not named in the statement, said.“Right now, there’s no way to ensure that our tools wouldn’t be leveraged to cause terrible harms or erode civil liberties away from public scrutiny. We’re talking about things like profiling individuals or targeting innocent civilians.”Google is one of several companies vying to fill the void left by AI start-up Anthropic and become the next go-to provider for government AI in classified and unclassified settings.Anthropic has sued the Pentagon over its designation of the firm as a “supply-chain risk” after the AI company requested that its technology not be used for mass surveillance in the US or for automated warfare.According to the letter organisers, Google has proposed contractual language that would prevent Gemini from being used for domestic mass surveillance or autonomous weapons without appropriate human control.The Pentagon, however, has pushed for broad “all lawful uses” wording that it argues is necessary to maintain operational flexibility.Employees said those proposed safeguards are technically unenforceable, pointing to Pentagon policy that prohibits outside entities from imposing controls on its AI systems.“If leadership is truly serious about preventing downstream harms, they must reject classified workloads entirely for now,” a second letter organiser said.Google already holds a contract with the US Department of Defence on non-classified workloads through a program known as genAI.mil, and the proposed new deal would extend Gemini’s capabilities into classified domains.The staff campaign draws direct inspiration from a 2018 employee movement that successfully pushed Google to abandon Project Maven, a Pentagon programme to integrate AI into drone operations.But in recent years Google has embarked on a strategy shift, steadily rebuilding its military business and competing with rivals Amazon Web Services and Microsoft for defence cloud contracts. AFPMore on this topicAnthropic says it won’t give US military unconditional AI useUS military relying on AI as tool to speed Iran operationsSee more onUnited StatesAI/artificial intelligenceGoogleDefence and military

This excerpt is published under fair use for community discussion. Read the full article at Straits Times — World.