Anthropic releases Claude for Government for US agencies

Anthropic has announced the general availability of ‘Claude for Government’ (C4G), a specialized AI platform designed for United States federal and state agencies. This release follows a public beta phase that began in July 2026 and includes tools such as Claude Code and Claude Cowork.

The platform is built to meet strict security requirements, operating within a FedRAMP-High authorized environment. It is specifically tailored to handle sensitive tasks, including intelligence analysis and threat assessments. To manage costs and access, Anthropic utilizes a usage-based billing model with prepaid budgets and mandatory spending caps, allowing administrators to distribute credits and set limits for different user groups.

Security features include single sign-on integration with agency identity services, audit protocols, and a ‘four-eyes principle’ for sensitive operations. While Anthropic faces competition from OpenAI in the federal market, the rollout continues despite previous regulatory challenges regarding model usage policies.