A promise to the next generation.
Three commitments we will never compromise — the line between a child who is used by AI and a child who leads it.
Safety you can inspect, not just trust.
A child product can’t hide behind a warm promise. Here’s exactly how Barnabut protects your child — in plain language, with the operational detail schools and parents actually need.
Inspect how safety works.
Every AI moment your child could meet passes through all five layers. Pick a moment, run it through the chamber, and watch each layer check it in turn.
Five promises. Each one you can inspect.
These are the five commitments parents ask about most. Each has a plain-language promise, a live status and a small proof you can try right here — then open its page to see exactly how it behaves and what happens when something goes wrong.
No child account begins without a verified adult. You create it, approve your child and stay in control.
No advertisements, sponsored recommendations, autoplay or endless feed — ever.
A child’s data belongs to the family. It never becomes an advertising product.
Governed technology plus trained human review, monitoring and escalation.
A guided learning space — not a social network. No unknown-adult contact.
What we collect — and what we never touch.
The rule is simple: the least we need to help your child learn, and nothing more.
- Your name and email — to set up the account and reach you
- Your child’s first name and age — to place them at the right stage
- Mission progress and growth signals — to help them learn
- What you choose to tell us — your hopes and worries, read by a real person
- A child’s last name, home address, or location
- A child’s photo — unless you choose to add one later
- Biometrics, contacts, or camera/mic access without consent
- Sell or rent data, show ads, or build ad profiles — ever
- Give a child an open-ended chatbot
Built for children. Governed for the world they are growing into.
Safety, privacy, family control and responsible AI are not settings added later. They shape how Barnabut works from the beginning. Select any protection to watch how it works — and what it blocks.
These describe how we design and operate — not formal certifications. Framework wording remains subject to legal review. Tap any item to see what the principle expects, how Barnabut addresses it, and what your family can inspect.
A child should never have to trade privacy, safety or trust to learn about AI.
The answers a promise can’t give.
Tap any question for the real, working answer.
You create the account and give explicit, informed consent before your child does anything. You approve your child’s profile, choose what’s shared, and can withdraw consent at any time. For children under the relevant age, a verified parent/guardian is always the account holder.
Inside a mission, the AI can help a child think: suggest, translate, give examples, and show its reasoning. It can never act as an open-ended friend, give medical, legal, or personal advice, collect personal details, or operate outside a teacher-designed task. A human is always above it.
Teachers design and lead every mission, review flagged interactions, and see each child’s work. Anything a child writes that trips a safety signal is routed to a human, not resolved silently by a machine.
Layered filters block violent, sexual, hateful, self-harm, and age-inappropriate content in both what a child could see and what they could enter. Young children have no free-text channel to an open AI at all.
No. We never show ads, never sell or rent data, and never build advertising profiles of children. Our model is subscriptions and school partnerships — not attention or data.
Younger children (6–11) get shorter, fully teacher-led sessions with no open AI input and no chat. Older students (11–15) get more independence with AI — always governed, always logged, and always with the ‘question it, check it’ habit built in.
Any parent or teacher can flag a concern from inside the product or by writing to our safety team. Child-safety reports are prioritised and reviewed by a real person, not a queue-bot.
You can delete your child’s data at any time, from your account or by asking us. Deletion is permanent and covers the learning record — no ‘soft delete’ that quietly keeps a copy.
We build to COPPA (US, under-13), GDPR-K (EU children’s data & parental consent), and keep FERPA-ready language for school deployments — and we align to UNICEF’s guidance on AI that upholds child rights, safety, and wellbeing.
Still have a question we haven’t answered?
A real person on our safety team will answer it — not a bot, not a form letter.
This is a plain-language summary. The full Privacy Policy, Child-Safety Policy, and Parent Promise are linked in the footer of every page.