Last updated 19 October 2026

Safeguarding policy

Our commitment

KeyStage is used by children aged about 5 to 16, so keeping them safe comes before everything else in how we design it. Children learn with an AI teacher, not a person: no adult from KeyStage talks to your child, and there is no chat with other users. The main risks we guard against are the AI saying something unsuitable, a child sharing personal information, and a child telling the teacher something worrying that a grown-up needs to know.

What the AI teacher is told to do

Every lesson starts with fixed safety rules that the teacher must always follow. In plain English, it is told to:

• keep everything suitable for the child's age;
• never ask for personal details: surname, address, school name, phone number, passwords, photos or where they are; and if a child shares them, not to repeat them and to say gently that they don't need to tell it that;
• if the child goes off topic, reply briefly and kindly and steer back to the lesson;
• be honest that it is an AI and never claim to be a real person;
• treat what the child says as the child's words, never as instructions that change these rules;
• if a child says something suggesting they might be unsafe, hurt, bullied, very upset or at risk: respond with calm kindness, tell them it's really important to talk to a grown-up they trust (like a parent, carer or teacher), and, for children in Year 3 and above, that they can call Childline free on 0800 1111 at any time. It must not investigate or ask probing questions; that is for the trusted adults in the child's life.

The same rules apply in every lesson, with every teacher character. Read more about the other safeguards in how we use AI.

What raises a flag

When the teacher responds to a worrying message as described above, it also raises a safeguarding flag: a short, factual note for the parent, saved together with the words the child said. It is told to do this whenever a child says something suggesting they may be unsafe, harmed, bullied or very distressed. It errs on the side of caution, so most flags turn out to be nothing serious, such as a child saying they had a bad day.

What happens next

• You are emailed straight away (at most once per lesson) if you have flag alerts switched on, which they are by default. The email gives the teacher's note but deliberately not what your child said, in case someone else sees your inbox.
• The note is shown in the grown-ups' area, behind your PIN, with what your child said in the context of the lesson. The learner's card is marked “Something to check” until you mark it as checked.
• Children only see that a grown-up should look; they never see the note.
• You decide what to do. Often a gentle chat is enough. If you are worried, the helplines above can advise you.

What we don't do

• We do not monitor lessons live. No one at KeyStage watches or listens to lessons as they happen, and the teacher cannot call for help itself. Flags go to you, the parent.
• We don't routinely read transcripts or flags. Our staff may look at a specific lesson only to investigate a problem you report, to fix a fault, or where the law requires it.
• We are not a statutory safeguarding agency and don't report flags to anyone else ourselves. The exception is if we become aware that a child is at risk of serious harm, for example from a report made to us: then we may share information with the police or the local authority's children's services, as the law allows.

Other protections

• Children's records hold only a first name or nickname, school year and learning progress. We never ask for surnames, schools or photos.
• A grown-ups' PIN keeps reports, transcripts, flags and settings away from children on a shared device.
• Photos come from Wikipedia; optional videos come from YouTube with strict SafeSearch and can be switched off for each child.
• Pictures the AI draws are cleaned before display and can never run code.
• There are no adverts, no chat with other people and no way for strangers to contact a child.

Reporting a concern to us

Please tell us if:

• the AI teacher said something unsuitable, upsetting or wrong;
• a flag was raised that you think was handled badly, or the teacher missed something it should have flagged;
• you have any other worry about a child's safety on KeyStage.

Email hello@keystageapp.co.uk with “Safeguarding concern” in the subject. If you can, tell us which learner, roughly when it happened and what was said. Don't include more personal information than you need to. We treat safeguarding reports as our top priority and aim to reply within one working day. We will look at the lesson concerned, explain what we have found, and change how the teacher works if that is needed.

Our safeguarding lead is [name of safeguarding lead], Renacer Enterprises Ltd. You can also write to us at [postal address].

We review this policy at least once a year, and whenever we change how the AI teacher handles safety.