The Pulse
Anthropic Makes Claude for Government Generally Available
Anthropic made Claude for Government generally available to federal and state agencies on September 30. The FedRAMP High-authorized service offers desktop Claude, capped usage billing and administrative controls, while Claude Code CLI and C
AI.info Team · Written with AI ·
Anthropic made Claude for Government generally available to federal and state agencies on September 30, ending a public beta that began in July. The offering brings Claude’s coding and agentic capabilities to a FedRAMP High-authorized environment, with agency controls for budgets, users and oversight. Anthropic says agencies can start without arranging a separate relationship with a cloud provider.
The service is designed for desktop work, including drafting memos, reviewing requests for proposals and processing casework. Public-sector teams can also use Claude to build or modernize software systems that support public services. Claude Code’s command-line interface and Claude for Microsoft 365 are not generally available yet; Anthropic says both are entering early access through the same environment.
Prepaid use replaces per-seat fees
Agencies pay for usage in fixed increments, with no seat fee and a hard cap tied to the amount the agency has obligated. Administrators can assign user tiers with spending and model limits, monitor usage by user and model, and receive alerts as a balance runs low. The structure gives departments a way to set spending limits before staff use the service rather than relying on a seat count alone.
Anthropic’s controls also reflect how government organizations divide responsibilities. Department administrators can distribute prepaid usage to sub-agencies, which manage their own users. Agencies can connect their identity provider for single sign-on, while SCIM group mappings set rate limits, dollar caps and permitted models for different user tiers.
Audit logs and local conversation history
The announcement describes an audit log that records administrative actions and can be reviewed by organization administrators. Anthropic says sensitive operations on its side require approval from two people. Usage exports contain metering data rather than sensitive material, while conversation history stays on the agency-managed device.
That separation matters for agencies documenting approvals and oversight: the company says exports can support authorization-to-operate and inspector-general requests without moving sensitive content. The application can be deployed through standard agency mobile device management platforms, and existing customers can import their conversation history into the desktop app.
Availability comes with a defined boundary
Anthropic says the generally available product is for federal and state agencies. The company’s announcement does not extend that availability to local governments, even as its broader government offerings have previously addressed other public-sector customers. Agencies can request access through Anthropic’s government sales channel, and the company says procurement officers can contract with it directly under general-availability terms.
The practical distinction is between the desktop product and the two early-access additions: Claude for Government is available now, while Claude Code CLI and Claude for Microsoft 365 remain limited to early access. For agencies weighing adoption, the published terms center on prepaid, capped usage, department-level administration and conversation history kept on agency-managed devices.