Google Negotiates Pentagon Deal to Deploy Gemini AI in Classified Settings
Google is negotiating with the Department of Defense to deploy its Gemini AI models in classified environments, proposing contract clauses to bar domestic mass surveillance and autonomous weapons without human oversight. It seeks parity with OpenAI’s Pentagon agreement while key uncertainties remain over finalizing surveillance and autonomy provisions.
1. Negotiation Overview
Google is engaged in high-level discussions with the Department of Defense to enable deployment of its Gemini AI models within classified networks, marking a strategic expansion of its role as a Pentagon technology contractor.
2. Proposed Safeguards
The company has drafted specific contract language to prohibit domestic mass surveillance and to ban any autonomous weapons targeting capabilities without direct human supervision.
3. Industry Alignment
These proposed clauses closely mirror those included in OpenAI’s existing Pentagon agreement, indicating a move toward standardized ethical and operational guardrails across leading AI vendors.
4. Outstanding Issues
It remains unclear whether the Pentagon will accept Google’s proposed guardrails, with final approval hinging on resolution of remaining disputes over surveillance restrictions and autonomy controls.