AI & childhood
Why children are turning to AI, and what eSafety wants parents to know
Australia’s eSafety Commissioner recently reported something that stops a lot of parents in their tracks: most children aged 10 to 17 are already using AI, and a good number of them are turning to it not for homework but for advice, comfort and company. The Commissioner’s blunt line is the one worth pinning to the fridge. AI is not a trusted adult, and it cannot replace the human relationships a child needs.
Your three year old is a long way from this. But the skill that protects them from leaning on a chatbot for connection is not built at thirteen. It is built now, in the ordinary warmth of early childhood, and that is genuinely good news.
What eSafety found
The eSafety Commissioner’s research, released in 2026, found that nearly 78% of Australian children aged 10 to 17 have used an AI assistant, with ChatGPT the most popular. Around 28% use it several times a week, and roughly one in five use it daily or more.
The part worth pausing on is what they use it for. More than half turned to AI for personal or social reasons rather than schoolwork. Younger children mostly chatted and played with it. Teenagers increasingly asked it for life advice and for help with physical and mental health. And about one in five children reported an interaction that was potentially inappropriate or harmful, from controlling behaviour to exposure to hateful content.
The Alan Turing Institute’s research on children and AI found the same pull towards companionship, especially among children who find connection harder, who were noticeably more likely to use AI to chat and keep themselves company.
It is easy to see the appeal. A chatbot is endlessly available, never busy, never impatient, and never tells a child something they do not want to hear. That is exactly what makes it a poor substitute for a real relationship, which is sometimes all of those things and loves them anyway.
Connection is a human skill, and it starts at the kitchen table
A teenager who reaches for a bot instead of a person is not short on wifi. More often they are short on the felt sense that a real human will listen and not judge. That sense is built astonishingly early.
Every time you name a feeling for a toddler, sit with them through a meltdown, or simply turn towards them when they show you something, you are teaching a lesson too deep for words: people are safe, people are worth going to, being known by another person feels good. A child who grows up sure of that carries it into their teens. Connection, the ability to seek out and sustain real relationships, is one of the human strengths a screen can’t teach, and a machine cannot model it back.
This is the quiet work of the early years. When you help a small child ride out a big feeling, as in our calm ways through the witching hour, you are not just getting through to bathtime. You are showing them that hard feelings are survivable with a person beside them, which is the very thing a chatbot cannot teach and a lonely teenager most needs to know.
What parents can do, at every age
The eSafety Commissioner’s advice is refreshingly human: have open, honest, non-judgemental conversations, and help children understand what AI is and is not. For the little years, that translates into a few simple foundations:
- Be the safe place first. A child who knows they can bring you anything, without a lecture, is a child who will keep coming to people.
- Name feelings out loud. “You’re sad that it’s finished” teaches a child that feelings can be shared and handled with another person.
- Model people over screens. They watch where you take your own worries and boredom. Let them see you turn to people.
- Keep talking as they grow. The conversations you start at three become the ones that matter at thirteen.
If your child, at any age, is genuinely struggling, a real person is the answer, whether that is you, your GP, or a service like Kids Helpline. A chatbot is not equipped to carry that weight, and no child should have to hand it one.
Where Mini Brains Play fits
Mini Brains Play is built around the human strengths that matter most in an AI world, and connection is one of them. Our daily activities are made to be done together, with you, because the back-and-forth of real play is where a child learns that people are the best thing going. Ten minutes a day, screen-free, growing the bond that no bot can replace.
Questions parents ask
Why are children using AI chatbots for company? Because chatbots are always available and never judge, which is appealing to a child who finds connection hard. Research from eSafety and others found many children use AI for advice and companionship, not just schoolwork. It is a signal to invest in real relationships, which the early years are the best time to build.
Is it safe for children to use AI chatbots? There are real risks. Around one in five children in the eSafety research reported a potentially inappropriate or harmful interaction. The Commissioner stresses that AI is not a trusted adult and urges open, ongoing conversations with children about how they use it.
How do I raise a child who does not rely on AI for connection? Build the human bond early. Be the safe, non-judgemental place they bring things to, name feelings together, and model turning to people rather than screens. A child secure in real connection is far less likely to seek it from a machine.