On Friday, the Pentagon announced it had struck deals with OpenAI, Google, Microsoft, Amazon, Nvidia, Elon Musk's xAI, and startup Reflection to deploy their artificial intelligence tools within classified military and defense environments. The agreements permit these companies to provide AI capabilities for sensitive government operations—a significant expansion of commercial AI into national security infrastructure. However, the announcement immediately drew attention for a notable absence: Anthropic, the San Francisco-based AI safety company founded by former OpenAI researchers, was not included in the partnerships. This omission raises questions about how the Defense Department evaluated competing AI vendors and what criteria guided the selection process.
The practical implications of these classified deals remain opaque. Defense officials have not publicly detailed whether the agreements grant companies code access to military systems, involve advisory relationships with defense planners, or constitute something else entirely. "Classified settings" can mean different things operationally—it could range from using commercial AI models for unclassified analysis of classified problems to direct system integration with defense networks. Without clarity on the scope and nature of these partnerships, it is difficult to assess how extensively these commercial AI systems will influence military decision-making or what security protocols govern their deployment. The Pentagon did not immediately respond to requests for comment on selection criteria or the specific operational parameters of these agreements.
Anthropic's exclusion is particularly noteworthy given the company's founding mission around AI safety and alignment—principles that might theoretically align with government interest in controllable, reliable AI systems. The company has previously engaged with policy discussions around AI regulation but has limited documented work with defense agencies compared to competitors. Market analysts note that Pentagon procurement decisions carry significant symbolic weight in the AI industry, potentially signaling to investors and customers which companies the U.S. government views as strategically viable partners. Whether Anthropic's safety-first positioning affected its exclusion, or whether the decision turned on other technical or commercial factors, remains unclear without official guidance from the Defense Department or the companies involved.