Over 580 Google Employees Urge Pichai to Refuse Classified Pentagon AI Work
More than 580 Google staff — including 20+ directors and senior DeepMind researchers — sent CEO Sundar Pichai a letter Monday. They demand Google refuse classified Pentagon AI work that would put Google models on air-gapped military networks.
More than 580 Google employees — including 20+ directors and VPs and senior DeepMind researchers — sent CEO Sundar Pichai a letter Monday urging the company to refuse classified AI contracts with the US Department of Defense. Reuters reported on April 16 that Google and DoD had been negotiating use of Google AI for "all lawful uses," including classified workloads.
The Letter
Coordinated by staff at DeepMind, the letter argues that on air-gapped classified networks, Google cannot meaningfully monitor how its models are used — meaning "trust us" becomes the only guardrail against autonomous weapons targeting and mass-surveillance applications. "We want to see AI benefit humanity; not to see it being used in inhumane or extremely harmful ways," the letter reads. As of Monday, Pichai had not publicly responded.
The Concern
This is the largest internal AI-ethics protest at Google since the 2018 Project Maven walkout, which forced the company to drop a Pentagon drone-imagery contract. Since then, Google has loosened its 2018 self-imposed ban on weapons-related AI work, joined the JWCC cloud contract, and pursued military and intelligence customers more aggressively. The DeepMind cohort signing this letter suggests the cultural divide between Google's research arm and its commercial cloud business has not healed.
The Washington PostThe Stakes
Two outcomes are possible and they matter for the whole industry. If Pichai refuses the classified work, Google effectively cedes the multibillion-dollar military AI category to Anthropic, OpenAI, Palantir, and Scale, all of whom have leaned in. If he greenlights it over employee objections, expect departures — particularly from DeepMind — and watch where those researchers land, because they take the alignment-and-safety culture with them.
Crypto + AI policy moves, the day they land — subscribe →
How we report: This article cites primary sources, regulatory filings, and on-chain data where available. BlockAI News uses AI tools to assist with research and first-draft generation; every article is reviewed and edited by a human editor before publication. Read our full How We Report page, Editorial Policy, AI Use Policy, and Corrections Policy.
