F. AI System
AI Proactive Behaviour
Built-in behaviours that keep conversations helpful: clarifying out-of-scope questions, offering an optional email follow-up, and guiding visitors.
By aninditoUpdated 6 Sep 2026
Privas AI may add a short follow-up message after an answer when more context or guidance would help. This keeps conversations clear and productive. Visitors can ignore these messages and keep chatting normally.
Built-in behaviours
Clarify when a question is outside the domain
If a visitor asks about something unrelated to the domain's knowledge, the assistant explains that the topic is outside its scope and guides them back to relevant topics. This prevents misleading answers.

Offer an optional email follow-up
During longer conversations, the assistant may ask if the visitor would like to share their email, so it can send a summary or follow-up after the chat. Sharing is always optional.

Guide the visitor
The assistant may suggest what a visitor can ask, such as topics it can explain or common next steps, without interrupting the conversation.
Going further with Proactive Rules
The behaviours above are built in. You can also define your own rules that show a component when a condition matches, for example a promo card when a visitor asks about pricing. See Proactive Rules.
Why this matters
Proactive behaviour improves the visitor experience: it prevents incorrect answers, guides visitors toward useful questions, and captures follow-up contact when appropriate. The assistant behaves less like a generic chatbot and more like a helpful representative.
Next step
To understand how the assistant retrieves information, continue to AI Knowledge Sources Explained.