Q&A: Anthropic's life sciences team talk Claude's first scientific discovery

Anthropic has announced the general availability of Claude for Government, a specialized version of its Claude AI model designed specifically for government agencies. Launched as a public beta in July 2026, Claude for Government operates within a FedRAMP High-authorized environment, ensuring robust security for government employees.

The new platform allows government workers to access the same tools as commercial Claude customers, while also providing additional controls such as model restrictions, usage monitoring, spending allocation, and management of external connections. Anthropic emphasizes that government agencies will not be left behind, as new features will be rolled out to Claude for Government at the same time as they are introduced to commercial versions of Claude.

In response to concerns about AI costs and public sector scrutiny, Anthropic has revamped its pricing structure for Claude for Government. Instead of per-seat billing, agencies can now commit to fixed increments, with usage limited to the allocated amount. Administrators can define user tiers with spend and model limits for each group, track usage by user and model, and receive alerts when balances are low.

Additionally, Anthropic is extending Claude's availability to Microsoft 365 via early access. This means that Claude tools will be integrated directly into Microsoft's Word, Excel, PowerPoint, and Outlook software. By committing to the same public cadence for new feature releases and including Claude for Microsoft 365, Anthropic acknowledges the importance of maintaining feature parity and security standards for government entities.