Two questions matter here and they pull in opposite directions: is my child safe, and am I watching too closely. This article answers both honestly — including the things we deliberately do not show you, and why.
Watch: Safety and what you can see
1 min 22 s · presented by Jason, Mandarin Temple's AI tutor
Captions are part of the picture, under Jason. The full transcript is below.
Read the transcript
- Hi, I’m Jason, Mandarin Temple’s AI tutor. Here’s what protects your child here, and what you can and cannot see, shown with a demonstration family.
- You see your child’s learning record, what has stuck and what has faded, but never the words they typed.
- On their own page, What stays theirs says it plainly: no answer, message or note they wrote reaches you there.
- Their results on the optional cognitive tasks stay theirs as well, shown to them alone, not to you or a teacher.
- Research use and those tasks stay off until you agree, and the server won’t take that agreement from your child’s own account.
- Under Activity, the page-by-page replay of their evening is left out for you, because it shows where they went, not what they learned.
- Private messages are limited on the server too: a child’s account can exchange them with nobody but you, their guardian.
- And their account stays out of public view: private circles only, off the public leaderboard, and never in the tutor directory.
- That’s what protects your child: you see how their learning is going, and only you can agree on their behalf.
Open the video guide on its own page, with its key moments
What you can see
- Progress — Lessons completed, and how far through each HSK level they are.
- Activity — When they last practised, how many days this week, their streak, XP and study time.
- Skills — On their own page: what they recognise against what they can produce, and how each skill is moving over time.
All of it is read the same way your child’s own screens read it, so you and they see the same figures. Your child is told, on their own Learning Lab page, what you can see.
What we do not show you
- Anything they typed — Not their conversations with the AI tutor, not their free-text answers, not their messages. No endpoint behind a parent view returns typed text.
- Their moment-to-moment behaviour — The detailed session-flow data available in a learner’s own Lab is withheld from guardians. You get outcomes; you do not get a replay.
- Their private practice as a transcript — You can see that they practised speaking and how they scored. You cannot read what they said.
Who can reach your child
People in the same circle, and nobody else. That is enforced on the server rather than hidden in the interface: a message to someone outside your circles is refused.
There is no site-wide directory of learners and no way to search for a person. A guardian-managed child account is also kept off public surfaces entirely — it does not appear in the tutor directory or any other public listing.
The AI tutor
The tutor’s instructions live on our server, not in the browser, and the topic is fenced — it is a Chinese-learning tutor and cannot be talked into being something else by a message from the page.
You will not see the conversation. What you will see is that it happened, and the progress that came from it.
Watching without hovering
The dashboard makes daily inspection possible and that is mostly a trap. Three things that work better:
Look weekly, not daily — a week is long enough to show a pattern and short enough to act on. Ask about the thing they did rather than the thing they missed. And when you do notice a stall, ask where they got stuck rather than why they stopped; the first is a question about a lesson and the second is a question about them.
Common questions
No. Nothing typed is exposed to a parent view. You see that practice happened and what came of it, not the transcript.