Supercharge regulated workloads with Claude Code and Amazon Bedrock
Amazon Web Services Bradley Wyman
AWS put Claude Code on Bedrock in GovCloud for regulated dev work. It gives agencies and other sensitive teams an AI coding path with compliance controls built in.
Based on reporting by Amazon Web Services, Bradley Wyman — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
AWS is pushing Claude Code into one of the few places where AI tooling has to behave like a grown-up: GovCloud. The company says Anthropic’s Claude Opus 5.5 and Claude Sonnet 5.5 are now available in AWS GovCloud (US) Regions, alongside Claude Sonnet 5, as an on-ramp for AI-assisted development in regulated environments, including ITAR workloads.
The pitch is not just “use a coding assistant,” but “use one without blowing up your compliance story.” AWS says Amazon Bedrock in GovCloud is built with security-focused features, including customer content that isn’t stored, logged, or used to train AWS models or shared with third parties. It also points to FedRAMP Class D and DoD authorization pathways, plus integration with existing security controls in AWS GovCloud.
There are two endpoint surfaces here, and AWS is pretty clear about when each one matters. The bedrock-runtime endpoint uses the AWS SDK and supports Guardrails, Knowledge Bases, Agents, and invocation logging, which makes it the safer default for teams that need audit trails. The bedrock-mantle endpoint speaks the Anthropic Messages API natively and exposes features like server-side tools, background inference, and Projects. Both work with Claude Opus 5.5, Claude Sonnet 5.5, and Claude Sonnet 5, though Mantle is only available in AWS GovCloud (US-West).
Claude Code itself is the other half of the story. It runs in the terminal, in editors like VS Code and JetBrains, and through the Claude Agent SDK. AWS says it can edit multiple files, run tests, search Git history, resolve merge conflicts, create commits and pull requests, and connect to tools like the AWS CLI, Terraform, and Kubernetes through MCP.
The rest of the post reads like an enterprise deployment checklist, which is probably the right tone. AWS walks through IAM Identity Center, environment variables, model pinning, per-user token guardrails, prompt caching, and rate limits. It also makes the tradeoff obvious: Opus 5.5 is the heavier model and costs more per token, while Sonnet 5.5 is the cheaper default for many teams, with Sonnet 5 called out for workloads needing IL4/IL5 authorization.
My take — AI-written commentary, not fact-checked reporting
This is the rare AI rollout that seems aimed at auditors before hype decks, which is a good sign. Most vendors sell “enterprise readiness” the way restaurants sell “artisan”; AWS at least hands over the forms, the guardrails, and the bill. The catch is still the same: putting a smart model in a compliant cloud does not make sloppy governance disappear.
Read more about this at: Amazon Web Services