The Pentagon finalised Pentagon AI agreements with seven technology companies on May 1, 2026, to deploy artificial intelligence tools on classified military networks.
The agreements include Google, Microsoft, Amazon Web Services, Nvidia, OpenAI, SpaceX and Reflection AI. The Defence Department’s use of seven vendors signals an effort to avoid relying on a single AI provider for classified systems.
Anthropic was excluded from the agreements after a dispute with the Pentagon over AI safety guardrails.
The company had demanded limits preventing military use of its technology for autonomous weapons or mass surveillance.
Anthropic’s Claude had previously been the only AI model available to the department through the classified network.
In March, the Pentagon labelled Anthropic a “supply chain risk” and banned its use across the Defence Department and its contractors.
Read: Google Pentagon AI Deal Covers Classified Networks
President Donald Trump also said his administration would cut ties with the company. Anthropic filed suit, and a federal judge recently halted the government’s move.
Pentagon personnel have six months to phase out Anthropic’s programs, though officials have been reluctant to move away from the tools because they view them as superior.
