Services
Summary
Microsoft Fabric will limit workspace roles to 1,000 users or groups from August 1 to August 31, 2025. Overlimit workspaces can’t add more users/groups until under the limit. Admins should consolidate access and use provided APIs to manage workspaces.
Details
Microsoft Fabric will implement workspace access limits from August 1 to August 31, 2025, as part of ongoing efforts to improve service quality and reliability. Once this limit is rolled out, each Fabric workspace will be limited to a maximum of 1,000 users or groups in workspace roles (Admin, Member, Contributor, Viewer). This change does not affect the number of users within a group itself. For example, a workspace with one group, that group is permitted to have over 1,000 users.
When this will happen:
This change will roll out between August 1, 2025, and August 31, 2025.
How this affects your organization:
Once this feature is rolled out:
- A maximum of 1,000 users or groups can have workspace roles for a Power BI or Fabric workspace.
- Workspaces that are already over the limit will remain overlimit, but no additional users or groups can be added to workspace roles until the workspace is under the limit.
- For overlimit workspaces, workspace admins will need to consolidate user/group access to stay under the limit.
- Workspace admins or members attempting to add users or groups after this limit is reached, will get an error in the workspace access flow.
What you can do to prepare:
Admins can take the following steps:
- Retrieve all workspaces in your org by using this API: Admin - List all workspaces
- Retrieve access details for each workspace using this API: Admin - List workspace access details.
- Identify workspaces that are near, at, or over the limit based on access details count in step #2
- Identify workspace admins for workspaces that are near, at, or over the limit from step #3 and step #2
- Work with workspace admins to consolidate access by (a) Grouping users by role (Admin, Member, Contributor, Viewer) (b) Assigning access via groups instead of individual users