Azure AI + machine learning built-in role

Cognitive Services LUIS Reader

Formerly provided LUIS read and test access through resource read Actions and LUIS DataActions. LUIS authoring and runtime endpoints are fully retired as of March 31, 2026, and requests fail. This overlay is retained only for migration and access cleanup; it is not a recommendation for a new assignment.

Control-plane actions (3)

• Microsoft.CognitiveServices/*/read
• Microsoft.Authorization/roleAssignments/read
• Microsoft.Authorization/roleDefinitions/read

Data-plane actions (2)

• Microsoft.CognitiveServices/accounts/LUIS/*/read
• Microsoft.CognitiveServices/accounts/LUIS/apps/testdatasets/write

Excluded actions (0)

None

Assignable scopes (1)

• /

Practical scope

Azure role assignments apply at the selected scope and are inherited by child scopes. Use the narrowest supported resource, resource-group, subscription, or management-group scope that contains the intended resources. Existing assignments for this retired service must be removed at the scope where they were created; an inherited assignment cannot be removed from a child resource.

Common use cases (2)

• Inventory principals and automation that still hold this legacy role before migration or cleanup begins. Export any recoverable LUIS application data and rebuild the workload in Conversational Language Understanding or a current Microsoft Foundry offering; select current roles for that replacement independently.
• Remove stale direct or inherited assignments after the replacement workload and rollback plan have been verified.

Prerequisites (2)

• Identify the legacy resource, every direct and inherited assignment, and the principal or group that received the role.
• The cleanup administrator needs Microsoft.Authorization/roleAssignments/delete at the scope where each assignment was created.

Best practices (4)

• Do not create a new assignment for this retired service.
• Export any recoverable LUIS application data and rebuild the workload in Conversational Language Understanding or a current Microsoft Foundry offering; select current roles for that replacement independently.
• Select roles for the replacement service from its current product documentation rather than carrying this legacy role forward.
• Remove assignments at their originating scopes after migration validation and retain an auditable record of the cleanup.

Security considerations (2)

• The legacy role exposed LUIS application content and test results even though it did not provide full lifecycle management.
• A stale parent-scope assignment remains inherited by child resources until it is removed at the originating scope.

Assignment guidance

Do not assign this role. Locate existing assignments in Access control (IAM), confirm the retired workload no longer depends on them, and remove each assignment at its originating scope. Reassess access separately for the replacement service.

Editorial sources (6)

• Azure built-in roles for AI + machine learning - Azure RBAC | Microsoft Learn →

  Supports: Description, Security considerations. Retrieved 2026-07-17.

• Understand scope for Azure RBAC | Microsoft Learn →

  Supports: Practical scope, Assignment guidance. Retrieved 2026-07-17.

• Best practices for Azure RBAC | Microsoft Learn →

  Supports: Best practices. Retrieved 2026-07-17.

• Understand Azure role definitions - Azure RBAC | Microsoft Learn →

  Supports: Security considerations. Retrieved 2026-07-17.

• Remove Azure role assignments - Azure RBAC | Microsoft Learn →

  Supports: Practical scope, Common use cases, Prerequisites, Best practices, Assignment guidance. Retrieved 2026-07-17.

• Migrate to Azure Language from Language Understanding (LUIS) or QnA Maker - Foundry Tools | Microsoft Learn →

  Supports: Description, Common use cases, Best practices, Security considerations, Assignment guidance. Retrieved 2026-07-17.

Official Microsoft Learn documentation →