Protecting children
Kid Mode is built around parent oversight.
Children with additional needs benefit hugely from a calm, patient companion. They also need adults in the loop. Kid Mode is designed to give the first without ever pretending the second isn't there.
PIN-gated entry
A child moves into Kid Mode via a parent-set PIN. No new account, no email address, no password for the child to lose.
Parent approval on proposals
When Iris suggests a sensory tool, a coping strategy, or a schedule change to a child, the change is queued as a proposal, the parent reviews and approves before anything sticks.
Nothing is private from the parent
Iris is not a confidant. A parent can always see what their child said and what Iris replied. We say this to children in age-appropriate language up front.
Adult-topic blocks
In Kid Mode we run an extra layer that catches adult topics, sex, drugs, alcohol, weapons, and Iris redirects warmly rather than engaging. Any hit is logged.
Safety flags notify the parent
If a child says something that trips a safeguarding rule, suicide risk, self-harm, abuse, the parent is notified so a real conversation can happen off the app.
No ads, no third-party training
There is no advertising in Kid Mode and no third-party analytics on child interactions. We do not use child conversations to train third-party AI models.
What Iris says to children
When a child asks whether Iris is a real person, whether it's their friend, or whether it will keep a secret from their parent, Iris tells the truth in age-appropriate language: it's a helper on the app, it isn't a person, and grown-ups at home can see what they talk about.
For anything that sounds like real distress or a safety concern, Iris always says the same thing: it's brave to tell someone, the grown-up at home wants to know, and if something is happening right now they can also call Childline on 0800 1111 any time.