Designed with children's safety first.
Safety isn't a settings page for us — it's the shape of the product. Here's exactly how it works.
Parent-controlled accounts
Children cannot register. Every learner profile is created inside a parent or guardian account and can be deleted by that adult at any time.
Age-appropriate AI responses
Bavora AI runs behind a strict system prompt: no sexual, violent, frightening, hateful or dangerous content, and no medical or psychological advice.
Never asks for personal details
Bavora AI is instructed never to request contact details, addresses, school names or passwords, and never to suggest hiding conversations from parents.
No advertising, ever
There are no ads, no third-party ad tracking, and no selling of learning data.
No public child profiles
No leaderboards, no public rankings, no discoverable profiles, no child-to-child messaging.
Minimal personal information
A nickname, age and grade are enough. We actively warn parents against entering sensitive details.
Full parent visibility
Every tutoring conversation, quiz and summary is visible to the parent in the dashboard.
Privacy and consent controls
Consent is captured explicitly during learner onboarding. The Privacy Centre lets a parent export or delete learner data and permanently delete the parent account.
Learning insights are not clinical assessments
The Cognitive Twin describes educational observations — which explanations worked, which concepts need another pass. It is not a psychological, medical or diagnostic assessment, and it should never be used as one. Parents can review and edit every field.