Security and identity: Azure AI Foundry

Scope

This is the security and identity page for the Azure AI Foundry target, the hosted-cloud target of ADR-0011. It is a map into the canonical documents, which are the source of truth. Compare all three targets on the Deployment targets hub.

Managed identity over service principal, Entra security groups for role assignment, secrets referenced by name from Key Vault, and content-safety policy applied per deployment.

Read this	For
Identity and security	The Well-Architected security posture.
ADR-0005	The governing identity decision.
ADR-0007	Content safety and responsible AI.
SPIKE-04	The research behind both.