Anthropic Launches Claude for US Civilian Agencies

Anthropic has officially launched Claude for Government, making its AI models available to United States federal and state agencies. The platform, which has been running in an open beta phase since July, operates within a FedRAMP High environment, representing the strictest cloud security standard mandated by the US government. This launch allows public sector practitioners to deploy Claude while adhering to rigorous federal data protection requirements.

The rollout specifically targets civilian agencies due to an ongoing conflict between Anthropic and the Pentagon. In March, the military labeled Anthropic a supply chain risk after the AI firm refused to lift its prohibitions on using its technology for mass surveillance and autonomous weapons. Although Donald Trump ordered all agencies to stop using Claude, the military reportedly utilized the model during the Iran war before eventually phasing it out. Legal battles have yielded mixed results: in late August, a San Francisco judge struck down one of the risk designations as unlawful retaliation, but in late September, a Washington appeals court upheld the other. While a recent Sunday dinner between CEO Dario Amodei and Trump suggests a potential easing of tensions, the military ban remains a significant hurdle.

For government IT administrators and developers, Claude for Government offers the same feature set as Anthropic's commercial enterprise tier. The pricing model is pay-as-you-go with fixed budget caps, allowing agencies to manage costs predictably. Administrative controls enable managers to set specific budgets and model access permissions on a per-department basis. To ensure compliance and security, the platform includes detailed audit logs and requires a two-person approval process for sensitive administrative actions. Crucially for data privacy, all chat data remains stored locally on agency devices rather than being transmitted externally.