AWS expands Claude Code to GovCloud regions for FedRAMP and DoD compliance

Amazon Web Services has made Anthropic's Claude Code available in AWS GovCloud (US) regions, extending the agentic coding tool to organizations with strict compliance requirements, including those governed by International Traffic in Arms Regulations (ITAR). The move means defense contractors, federal agencies, and regulated industries can now use Claude Code alongside Claude Opus 5.5 and Claude Sonnet 5.5 on Amazon Bedrock without leaving government-authorized cloud environments.

Both Claude Opus 5.5 and Claude Sonnet 5.5 hold FedRAMP Class D certification. Claude Sonnet 5 meets Department of Defense Impact Level 4 and 5 requirements, making it the designated model for DoD IL4 and IL5 workloads. The deployment architecture splits across two endpoint surfaces to balance governance with developer flexibility.

Two endpoints, two different jobs

The bedrock-runtime endpoint handles audit trails and guardrails - the compliance scaffolding that regulated teams require. The bedrock-mantle endpoint provides direct access to native Anthropic APIs, including server-side tools and background inference capabilities. Developers get the full Claude Code experience without sacrificing the logging and controls their security teams demand.

Claude Code integrates with IDEs such as VS Code and JetBrains, automates CI/CD pipelines, and connects to external tools through the Model Context Protocol. The tool functions as an agentic coding assistant - it can reason through multi-step programming tasks, execute commands, and interact with codebases autonomously within defined boundaries.

Model selection and cost governance

AWS recommends Claude Sonnet 5.5 for teams prioritizing cost efficiency and Claude Opus 5.5 for complex, high-stakes coding tasks. Claude Sonnet 5 remains reserved exclusively for DoD IL4 and IL5 environments. Organizations can enforce per-user token limits, enable prompt caching to reduce redundant processing costs, and pin specific model versions to prevent unexpected pricing changes when newer models roll out.

The configuration reflects a growing reality in government IT: development teams want the same AI-assisted coding tools their private-sector peers use, but procurement and security requirements demand guardrails that commercial cloud regions don't provide. By bringing Claude Code into GovCloud, AWS lets agencies adopt the tool within existing compliance frameworks rather than forcing workarounds.

Why this matters for government and defense developers

For development teams operating under FedRAMP or DoD Impact Level requirements, this removes a friction point that has slowed AI adoption. Instead of maintaining separate environments for compliance workloads and AI-assisted development, teams can now work entirely within GovCloud. The two-endpoint model means security officers see the audit trails they need, while developers access the full Anthropic API surface. For IT managers, the token limits and model pinning provide budget predictability - a non-negotiable in government contracting where cost overruns trigger reporting requirements. Teams building or maintaining applications in regulated environments can evaluate these capabilities through AI Code Generation Courses that cover agentic coding tools and compliance-aware deployment patterns.