DE
← Other Athentra features

Athentra AI · Governance

Permission-Scoped AI Access

AI features act as the person using them, so results are scoped to that individual's own permissions; there is no back door around access control. Assistant conversations are kept for 24 hours by default and can be cleared at any time, and customer data is not used to train models. The assistant cannot become a route around the permission model.

  1. 01

    Pain Point

    What's the problem?

    An assistant that queries with elevated rights becomes a way to read data the asker was never entitled to see.

  2. 02

    Remedy

    How do we attack it?

    Have agents call each service as the person asking, with that person's own permissions

  3. 03

    Your Value

    What do you get out of it?

    An answer can never reveal more than the person asking could have found themselves