Pentagon signs AI deals with 8 companies for classified military use, excludes Anthropic

The Pentagon announced agreements on May 1, 2026, with eight AI companies — SpaceX, OpenAI, Google, Nvidia, Reflection AI, Microsoft, Oracle, and Amazon Web Services — to deploy their AI tools in classified military networks (Impact Level 6 and 7). The agreements allow the Defense Department to use these companies' technology for 'any lawful use' in secret and top-secret environments, expanding beyond the existing GenAI.mil platform used by 1.3 million Defense Department personnel. Anthropic was explicitly excluded after refusing to agree to unconditional use clauses, and the company's Mythos cybersecurity model has raised concerns in government and banking about its vulnerability-discovery capabilities. The Pentagon reportedly signed with competitors in part to pressure Anthropic back to negotiations. Reflection AI, a lesser-known startup that raised $2 billion in October 2025 and is backed by 1789 Capital (in which Donald Trump Jr. is an investor), is among the new agreements.