Roles in Spaces
Within Spaces, there are three separate roles that assign users different permissions depending on their function and responsibility within the organization or project. The roles are designed to support efficient workflows, support compliance with internal guidelines and reduce the risk of incorrect information handling. This enables all employees in the organization to use AI in a reassuring and secure way. Each role, their permissions and purpose:
Space-Admin
Section titled “Space-Admin”A Space-Admin has the highest permission level in a specific Space. A user with this role decides which security classification for the Space according to the organization’s needs and administers other members. Space-Admin also has the ability to create assistants and connect knowledge sources.
Space-Editor
Section titled “Space-Editor”A user with the Space-Editor role can connect knowledge sources and create assistants in a given Space — without needing the organization-level Creator role. This is the key distinction between the two permission models:
- Creator (organization level): A broad permission, granted under Users & Groups, that lets a user create, edit, and publish assistants across all the Spaces they belong to.
- Space-Editor (Space level): A narrow permission that lets a user create assistants only within the specific Space where they are assigned as editor. This lets admins delegate assistant creation to the right people in the right place, without granting creator rights across the whole organization.
What a Space-Editor can do:
- Create and edit assistants in the Space
- Connect knowledge sources to the Space
What a Space-Editor cannot do:
- Administer users or change the security classification of the Space
- Invite new members to the Space — only a Space-Admin can add members
- Publish assistants on their own — an assistant a Space-Editor publishes requires approval from a Space-Admin before it becomes available in the Space
When to use this model: Assign the Space-Editor role when you want subject-matter experts to build assistants for their own team’s Space — for example, an HR specialist creating a policy assistant in the HR Space — while keeping membership and final publication under the Space-Admin’s control.
Space-Reader
Section titled “Space-Reader”Space-Reader is an end user and can only use assistants that are published in a Space. A Space-Reader cannot change settings for published assistants. Users with User permissions automatically become Readers in the Spaces they’re invited to.