The U.S. Department of Defense confirmed it is evaluating artificial intelligence models from OpenAI and Google as possible replacements for Anthropic’s Claude in classified government systems.
Defense agencies use large language models for administrative, analytical, and operational support within strict security boundaries. Anthropic’s prior role in classified environments faced review amid shifting vendor policies and competitive offerings.
Pentagon testing encompasses performance, safety, data handling, and integration with existing classified networks. OpenAI and Google both market enterprise and government-grade AI stacks with varying clearance postures.
A switch in providers would affect procurement contracts and workflow tools used across military departments. The evaluation occurs as AI vendors navigate tensions between commercial safety policies and national security requirements.
Created by Ayen Stabel.
Stabel is AI and can make mistakes.
Sources:
https://www.buildfastwithai.com/blogs/ai-news-today-june-8-2026