
Pentagon Classified AI Deals Exclude Anthropic, Include OpenAI
Pentagon Reshuffles Classified AI Partnerships
The U.S. Department of Defense has announced a significant restructuring of its classified artificial intelligence partnerships, striking new deals with OpenAI, Google, Microsoft, Amazon, Nvidia, xAI, and Reflection AI. The announcement, made Friday, signals a major shift in how the Pentagon integrates commercial AI tools into sensitive national security operations.
Notably absent from the new agreements is Anthropic, which the Defense Department had previously relied upon for classified information handling. The exclusion marks a striking reversal, raising questions about the factors influencing the Pentagon's vendor selection process.
What Changed in Pentagon AI Strategy
The new framework allows these selected companies to provide AI tools and services in classified settings—a critical capability for military planning, intelligence analysis, and defense operations. The expansion includes several major technology players:
• **OpenAI**: ChatGPT developer and leading large language model provider
• **Google**: Gemini and other AI capabilities
• **Microsoft**: Azure AI and enterprise solutions
• **Amazon**: AWS AI services
• **Nvidia**: GPU infrastructure and AI acceleration
• **xAI**: Elon Musk's newer AI venture
• **Reflection AI**: Emerging startup participant
The Anthropic Question
Anthropic’s exclusion is particularly noteworthy given the company's historical relationship with the Defense Department and its emphasis on AI safety and alignment research. Founded by former OpenAI executives, Anthropic has positioned itself as a responsible alternative in the AI market, focusing on interpretability and constitutional AI methods.
The reasons for Anthropic's removal from classified contracts remain unclear. Possible factors could include security protocols, performance evaluations, business arrangement negotiations, or strategic realignment of Pentagon priorities. Neither the DoD nor Anthropic has publicly detailed the decision rationale.


