Anthropic Launches Claude for Government for US Civilian Agencies

On October 1, 2026, The Decoder reported that Anthropic has launched Claude for Government for US federal and state civilian agencies. The platform runs in a FedRAMP High environment, which the report describes as the strictest US security level for cloud services. Agencies using it get the same features as Anthropic's enterprise customers, pay only for what they use under fixed spending caps, and can set separate budgets for individual departments. The report also notes that the Pentagon will not use the platform, because it still classifies Anthropic as a supply chain risk. The source excerpt ends at that point, so no further detail is given about the basis for that classification, nor about pricing, a rollout timeline, or how Claude for Government differs from other government-facing offerings. Based on what this single report makes checkable, the matter currently rests at the launch stage: Anthropic has opened the platform to US federal and state civilian agencies on a pay-as-you-go model with fixed spending caps and department-level budgets, while the Pentagon is holding off on the grounds of its supply chain risk designation.