Can AI Be Turned Off or Limited to Certain Users if Requested? 

Not really. But in Requests, every use of AI is logged and fully auditable. Any time a Scope is saved that has AI-generated content inserted into it, that is permanently logged in the request history log. So, if agencies want to have a policy of "nobody should use this," they would at least easily be able to find violators of that policy.

Q&A AI- AI is used only to group questions by topic so the user can focus on all related questions at once. They can remove questions from the grouping or create their own groups. It does not answer any questions. They shouldn't have an issue with that, should anyone use the AI function.