Full Breakdown
Google Employees Urge CEO to Reject Classified Military AI Contract
4/28/2026, 9:52:10 AM
Proposed Pentagon Deal for Gemini AI
On April 27, more than 600 Google workers submitted an open letter to CEO Sundar Pichai demanding that the company refuse a proposed Department of Defense agreement that would place its Gemini AI model on classified networks. The deal would extend Gemini—currently used in the unclassified genAI.mil program—into “air-gapped” environments for classified missions, a move the employees say lacks transparent oversight.
Historical Context: Prior Military Contracts and AI Principles
Google’s military involvement dates to the 2018 Project Maven protest, which forced the firm to abandon a Pentagon contract that used AI to analyze drone footage. In response, Google adopted AI principles that barred weapon and surveillance uses. Last year the company removed the explicit weapons clause, prompting renewed scrutiny as it seeks new defense contracts alongside rivals Amazon Web Services and Microsoft.
Scale of Employee Opposition
Signatories include more than 20 directors, senior directors and vice-presidents, with roughly two-thirds identified publicly and the remainder anonymous. Reports vary: one source cites “more than 600” signatories, another notes “about 580,” and a third mentions “560” employees. About 40 % work in the AI division, a similar share in Cloud, and the rest are spread across other units.
Pentagon Position and Official Statements
Under-Secretary of Defense for Research and Engineering Emil Michael told Bloomberg that the Pentagon will first deploy Gemini agents in unclassified settings “then we’ll get to classified and top secret.” The department also seeks “all lawful uses” of Google’s AI tools, arguing that commercial firms should not dictate usage policies during wartime. Google’s spokesperson has not responded to requests for comment.
Employee Criticism and Ethical Concerns
Workers argue that classified workloads are inherently opaque, preventing any meaningful oversight of how Gemini might be employed. They warn that “Agentic AI is particularly concerning because of the level of independence it can get to. It’s like giving away a very powerful tool at the same time as giving up on any kind of control on its usage.” Concerns extend to lethal autonomous weapons, mass surveillance, and the inability to enforce safeguards on air-gapped systems.
Conflicting Figures and Information Gaps
The exact number of employees backing the letter differs across reports (560 – > 600). Details of the proposed contract language remain unclear; employees claim the company’s draft would prohibit domestic mass surveillance and autonomous weapons, yet the Pentagon pushes for broad “all lawful uses” wording. No public document outlines the specific technical or legal safeguards that would govern classified deployment.
Verbatim Quotes
- “Classified workloads are by definition opaque,” — Organising employee, Google (unnamed)
- “We are Google employees who are deeply concerned about ongoing negotiations between Google and the US Department of Defense,” — Letter, signed by Google staff
- “As people working on AI, we know that these systems can centralize power and that they do make mistakes.” — Letter, signed by Google staff
- “We feel that our proximity to this technology creates a responsibility to highlight and prevent its most unethical and dangerous uses.” — Letter, signed by Google staff
- “If leadership is truly serious about preventing downstream harms, they must reject classified workloads entirely for now,” — Organising employee, Google (unnamed)
Outlook and Next Steps
The open letter is slated for delivery to Pichai on Monday, but the company has not issued a response. Ongoing negotiations between Google and the Pentagon could shape the future of classified AI use, while employee dissent may influence corporate policy, public perception, and the broader tech industry’s approach to defense contracts.
