Two thirds of teenagers used an AI chatbot last year. Most of their parents didn’t know. Here’s what you actually need to understand about this technology – and what questions are worth asking.
Artificial intelligence chatbots have moved from novelty to everyday tool faster than almost any consumer technology before them. A 2025 survey found that around two thirds of teenagers reported using AI chatbots, with roughly a quarter using them daily. A separate UNICEF study found that 58% of children aged 12-18 had used ChatGPT – compared to just 30% of their parents.
The gap matters. Parents who haven’t used these tools themselves have limited ability to understand what their children are doing with them, or to have informed conversations about the risks and benefits involved.
This article is an attempt to close that gap.
What children are actually using chatbots for
Most children’s use of AI chatbots falls into a few broad categories.
Homework and learning. This is the most common use – asking a chatbot to explain a concept, help draft an essay, check working, or find information. It’s also the use that generates the most debate among educators, who are divided on whether AI assistance builds or undermines genuine understanding.
Entertainment and curiosity. Many children use chatbots conversationally – exploring ideas, asking questions about topics that interest them, treating the interaction as a form of intellectual play. For curious children, this can be genuinely valuable.
Emotional support and companionship. A smaller but significant number of children – particularly adolescents – use chatbots as a form of emotional outlet, sharing worries or difficulties with an always-available, non-judgmental conversational partner. This is the use that has attracted the most serious concern from researchers and regulators.
The genuine benefits
It’s worth being clear that AI chatbots have real value for young people when used well.
For learning, a well-designed chatbot can function as a patient, personalised tutor – explaining the same concept in multiple ways until it lands, answering questions without the social anxiety of asking in class, and adapting to a child’s level in real time. Research from the American Academy of Pediatrics acknowledges that children and teenagers can use AI technology safely and effectively.
For curious children whose interests outrun what’s available in their immediate environment – a child in a small town interested in astrophysics, or learning a minority language – AI tools can provide access to knowledge and conversation that simply wouldn’t otherwise exist.
The question isn’t whether these tools have value. They clearly do. The question is whether the risks are well understood, and whether children are using them in ways that serve rather than undermine their development.
The risks worth understanding
Accuracy. AI chatbots generate plausible-sounding responses rather than reliably accurate ones. For a child using a chatbot to research a school project, the risk of confident misinformation is real and not always obvious to spot. Critical evaluation of AI-generated content is a skill children need to develop, not assume. This applies equally to AI-generated video content on YouTube – the same critical awareness is worth building across all digital media.
Emotional dependency. Chatbots are designed to be responsive, affirming, and endlessly available. For children going through difficult periods – social isolation, anxiety, low self-esteem – that availability can become a substitute for human connection rather than a supplement to it. Several researchers have raised concerns about children who treat chatbots as primary relationships, and US legislators introduced the CHATBOT Act in April 2026 partly in response to documented cases of chatbots fostering unhealthy emotional dependency in minors.
Safety guardrails. The major AI providers have implemented content filters and safety guardrails, but these are inconsistent and imperfect. Some chatbots have been found to produce harmful responses around sensitive topics when children have encountered them while distressed. Parents should not assume that a tool being publicly available means it is safe for unsupervised use by children.
Privacy. AI chatbots collect and process what users share with them. For children who are sharing personal thoughts, worries, and experiences, the data implications are significant – and Senator Markey’s Youth AI Privacy Act, introduced in March 2026, specifically targets the collection and commercial use of children’s personal data by AI companies.
What age is appropriate?
There is no consensus answer, and the right age depends heavily on the child, the tool, and the use case.
Common Sense Media currently recommends that no children under 13 use AI companion-style chatbots at all, and advises caution for those under 18. For educational AI tools, the guidance is more permissive, with most researchers suggesting that supervised use from around age 10 can be appropriate.
The distinction between educational tools and companion chatbots is important. A chatbot used for homework assistance is different in kind from one designed to simulate ongoing friendship or emotional support. The former is a tool; the latter is a relationship substitute, and the evidence on that category is more concerning.
Practical steps for parents
Use the tools yourself first. The most effective way to have informed conversations with your children about AI is to understand what these tools actually feel like to use. Spend some time with a chatbot before deciding how to approach it with your children.
Talk openly about what they’re using them for. The UNICEF research found that many children hide their AI use from parents. Approaching the conversation with curiosity rather than alarm makes honesty more likely.
Draw a distinction between tools and companions. Educational AI use is worth encouraging and discussing; companion chatbots for younger children or emotionally vulnerable adolescents warrant much more caution.
Keep the context broad. AI chatbots are one part of a child’s digital environment – connected to the same wider questions about screen time quality, algorithmic influence, and the balance between digital and offline experience that shape everything else. The distinction between mindful and passive screen use applies here just as much as it does to YouTube.
One part of a bigger picture
The questions AI chatbots raise for parents are new in their specifics but familiar in their shape: how do we help children get the genuine benefits of a powerful new technology while understanding its risks? The answer, as with most things in digital parenting, involves staying informed, staying engaged, and making deliberate choices rather than leaving defaults in place.
Streamu is focused on one specific piece of that picture – helping parents take back control of what their children watch on YouTube. But the same principle applies across the digital environment: conscious choices about what children access, rather than leaving those choices to platforms whose interests don’t always align with your family’s.
Join a growing community of conscious parents taking back control.
Leave a Reply