Google DeepMind Staff Pressures Sundar Pichai to Reject Pentagon's Classified AI Workloads

More than 600 Google employees have signed an internal letter calling on CEO Sundar Pichai to formally prohibit the Pentagon from using the company's AI models for classified military purposes, according to reporting by The Washington Post. The move signals deepening tension between the tech giant's workforce and its leadership over defense-sector AI partnerships, a fault line that has repeatedly surfaced since the 2018 Project Maven controversy forced Google to abandon a similar Pentagon contract.

Organizers of the letter claim that signatories include employees from Google's DeepMind AI laboratory, along with more than 20 individuals holding principal, director, and vice president positions—representing an unusually high concentration of senior technical staff. The letter reportedly states that "The only way to guarantee that Google does not become associated with such harms is to reject any classified workloads. Otherwise, such uses may occur without our knowledge or the power to stop them." The language reflects ongoing employee concerns that classified defense applications could proceed without transparency or meaningful oversight, even within a company that has publicly committed to AI safety principles.

The internal pressure arrives amid a broader reckoning across the AI industry over military applications. Anthropic, a rival AI developer, is currently engaged in separate legal proceedings with the Pentagon over its own defense-related contracts. Google has not publicly confirmed whether it has received or is negotiating new classified-work requests from the Pentagon. The outcome of this internal dispute could shape whether Google pursues a different stance on government AI partnerships than competitors who have accepted such work, potentially affecting its positioning in a market where defense agencies are increasingly seeking advanced AI capabilities.