The Pentagon framed the agreements as part of a transformation into "an AI-first warfighting force." Eight frontier AI companies will plug their advanced capabilities into the Department of Defense's classified networks.
Scope of the Deals
Microsoft, Amazon and Google — alongside five undisclosed AI firms — will provide model inference, agentic workflows and decision-support systems inside classified infrastructure. The move accelerates operational validation of LLM-based and agentic engineering systems in the defence sector.
Anthropic's Restraint
Anthropic is not part of the deals. The company refused to grant the Pentagon unrestricted access to Claude for use in fully autonomous weapons systems and broad domestic surveillance. CEO Dario Amodei met with the White House on April 17; Trump said afterward a deal was "possible," but no binding agreement is in place yet.
The Military Future of AI
The development cements AI's strategic value to defence — and underscores the growing tension between technology companies' ethical principles and national security requirements. The industry is splitting into two camps: "integrated by default" vs. clear "red lines."
Cite this: AI Mevzuları · May 1, 2026 · aimevzulari.com/haberler/pentagon-microsoft-amazon-google-askeri-ai-anlasmasi