Security Classification

A user with the Space-Admin role has permission to decide which security classification should apply to a Space. Security classification controls which AI models and thus tools are available within a specific Space. A classification can thus ensure that sensitive data is only processed by models that run within Sweden or the EU. Which security classifications are available is administered at the organization level.

The security class functions as a filter. If a Space is classified as “High security (Sweden)”, all models that are operated outside Sweden will be grayed out or invisible when you create assistants there.

• This protects you from accidentally sending sensitive data to the wrong place.
• The classification is set under Settings for the Space (by Space Admin).

Security is paramount. The video clip below shows how security classification works in practice: from Admin creating the classes to you applying them to a Space. You’ll see how the system automatically locks unauthorized models to protect your data.