In a major development for AI's role in national defense, the Pentagon announced on Friday that it has secured classified AI deals with seven technology companies: OpenAI, Google, Microsoft, Amazon, Nvidia, Elon Musk's xAI, and the startup Reflection AI. These agreements grant the Defense Department access to deploy these companies' AI tools in classified settings—a critical capability as the military increasingly integrates artificial intelligence into strategic operations. The announcement represents a significant expansion of the Pentagon's AI infrastructure partnerships, positioning these vendors as essential to U.S. defense capabilities. However, the contracts' specific values, timelines, and the particular classified programs they support remain undisclosed, as is standard for defense agreements involving sensitive national security applications.
The most notable aspect of the Pentagon's announcement is what it omits: Anthropic, the AI safety-focused company behind the Claude chatbot and one of the largest AI laboratories globally, did not receive a classified contract. While the Pentagon has not publicly stated its reasoning for excluding Anthropic, industry observers have noted the company's historical emphasis on constitutional AI and safety guardrails—approaches that may not align seamlessly with defense department requirements for rapid deployment and operational flexibility. The exclusion is particularly significant given Anthropic's prominence in the AI sector and its substantial funding rounds. Meanwhile, xAI and Reflection AI, less established players, secured Pentagon access, suggesting the defense department prioritized factors beyond market share or funding levels when evaluating these partnerships.
The Pentagon's contract awards underscore the accelerating militarization of commercial AI capabilities and raise questions about how defense requirements will shape AI development priorities. With access to classified systems, companies like OpenAI and Google gain unprecedented influence over military applications, while Anthropic's exclusion may signal that the Pentagon favors vendors willing to prioritize operational capability over cautious safety-first approaches. These partnerships will likely reshape how AI companies balance commercial, research, and defense objectives in coming years.