The AWS Machine Learning Blog describes how to configure secure, multi-environment access to Claude Platform on AWS using a single subscription. The approach covers three access patterns: cross-account SigV4 for AWS workloads, workspace-scoped API keys for developers, and OIDC federation for external environments. Isolation is maintained at the workspace level within a dedicated AI Services account.
Why it matters
Organizations that need to serve different types of consumers—internal AWS workloads, developers, and external environments—from one subscription can use these patterns to manage access while keeping workspaces isolated.
Who should care
Cloud architects, platform engineers, and teams operating Claude Platform on AWS who need to structure access and isolation across multiple environments.