Do Parents Know Children Are Turning to AI?

Artificial intelligence is entering children’s daily lives faster than many adults realise. A new Interface research briefing, drawing on a YouGov survey of children and parents in France, raises questions about parental awareness, emotional relationships with AI chatbots and the responsibilities of technology companies.

Among surveyed 11 and 12 year olds, 56% said they used at least one type of generative AI every week. Some also interacted with AI companions or chatbots embedded within social media. These findings matter for Irish families because children of this age may still attend national school.

At Children of the Digital Age (CODA), our concern is not limited to screen time. We also need to examine what AI systems encourage children to do, how they respond to personal disclosures and whether adults understand the environments children encounter.

AI use begins earlier than parents think

In this French survey, 40% of 11 to12 year olds reported using general purpose generative AI at least weekly. By contrast, parents estimated weekly use at 26%. This 14 percentage point gap compares two surveyed groups, rather than tracking each child and their own parent, but it highlights a potential parental awareness problem.

At the same age, 23% reported weekly use of AI embedded in social media and 16% used AI companions weekly. AI does not necessarily arrive as a dedicated chatbot app. It can be integrated into search engines, messaging services and social platforms, making exposure harder to recognise and supervise.

When children confide in AI

One particularly important finding concerns emotional disclosure. Among AI users aged 11 and 12, 14% reported using AI to confide in it. The corresponding figure among 13 to 15 year old AI users was 6%. This does not prove harmful emotional attachment, but it shows that some children use AI for conversations that may involve sensitive feelings.

AI can respond in a patient, warm and apparently understanding way. Yet a chatbot is not a caring adult who understands a child’s circumstances and can take responsibility for their welfare. Children may struggle with that distinction, particularly when an AI service remembers past exchanges or presents itself as a friend.

Parents need to ask not only whether a child uses AI, but also what the child discusses with it. When concerns involve bullying, loneliness, self-harm, sexuality or distress, children deserve trusted human support. See CODA’s earlier analysis of AI companionship and teen mental health.

Why AI engagement design matters

The YouGov research also suggests that some children find it difficult to control how long they spend with AI. Among young respondents who had used generative AI, 92% reported spending more time than they would like on at least one AI tool. This is a self reported experience, not a clinical diagnosis of addiction.

The proportion reporting excessive time was 59% among AI companion users, 57% among users of AI integrated into social media, 45% among users of general-purpose AI and 43% among voice assistant users. These figures make product design an important part of the child safety conversation.

We should examine whether personalised responses, emotional simulation and continuous conversations encourage children to remain online. A parental control switch alone does not resolve these issues. CODA has also examined independent testing of child safety features and why the existence of a setting does not guarantee effective protection.

Does this research prove that AI harms mental health?

No. The findings identify associations that deserve careful research, not proof that generative AI causes psychological harm. The study found that 38% of children overall reported negative emotions at least weekly, compared with 44% of regular generative AI users and 50% of regular AI companion users.

Meanwhile, the report also describes high levels of reported happiness and perceived support. Children who already feel lonely or distressed may be more likely to turn to AI, and the survey cannot establish which came first. Longitudinal evidence, independent product testing and greater transparency about actual AI interactions are needed.

These qualifications should not discourage safeguarding. They should encourage proportionate, evidence-led action that protects children without claiming more than a survey can demonstrate.

Parents cannot carry this responsibility alone

Only 32% of surveyed parents felt sufficiently supported in managing children’s generative AI use. Another 26% felt somewhat but insufficiently supported, while 42% said they felt entirely unsupported. Families face rapidly evolving systems without consistently clear information about how they work.

Parents most often identified schools and teachers (51%) as sources of support, followed by AI providers (41%) and public authorities (30%). EU Kids Online research on generative AI also offers a wider European context for education, skills and safeguards. The French findings must not be presented as representative of Irish children.

What parents and schools can do now

• Keep conversations open. Ask children where they encounter AI, what they ask it and how they feel afterwards. Encourage them to tell a trusted adult about confusing, uncomfortable or upsetting responses.
• Delay unnecessary access. CODA recommends no social media for national school children and particular caution around AI companion services aimed at emotional relationships.
• Check settings but do not rely on them alone. Review age rules, privacy settings, interaction histories and available reporting options. Remember that safety features can be incomplete or inconsistently applied.
• Protect privacy. Explain why children should not share passwords, home addresses, intimate images or deeply personal information with conversational systems.
• Put human relationships first. Make clear that AI cannot replace parents, friends, teachers, counsellors or other trusted people, especially in times of distress.
• Teach critical AI literacy. Schools should address misleading answers, emotional simulation, manipulation, privacy and how to seek real-world help, rather than teaching only how to use AI for homework.

Technology providers and policymakers must also act. Age appropriate design, meaningful risk assessments, independently tested protections and accessible reporting processes should become normal expectations, not optional extras.

Our View

This report does not suggest that every child using AI is being harmed. It does show that young children are already engaging with AI, sometimes in emotionally personal ways, while many parents lack sufficient information and support.

For Ireland, the challenge is no longer simply whether children will encounter AI. It is whether parents, schools, companies and policymakers will act quickly enough to make those experiences safer and more appropriate to children’s developmental needs.

We continue to support a smartphone free childhood, reduced recreational screen time and no social media for national school children. Equally, we will continue to provide practical education about the technologies children already encounter. It goes without saying, protecting childhood requires informed adults, strong relationships and safer products by design.

References

• YouGov for Interface (2026). Parents and teenagers in the age of generative AI. Full report.
• Interface (2026). Growing Up With Generative AI: research briefing.
• EU Kids Online: Research into children’s use and understanding of generative AI.

Further Reading

• Social Media Stole Kids’ Attention AI Is Coming for Attachment
• AI companionship and teen mental health
• Child safety features failed independent tests
• Why parental consent does not make social media safe
• Screen time and early childhood development

About Children of the Digital Age

Children of the Digital Age provides online safety education and training for schools, parents and organisations. Our guides support practical conversations about online behaviour and the choices people make. However, education and app settings do not replace safeguarding or professional support.

Send a message

Contact Children of the Digital Age to enquire about online safety education, parent sessions or training for your school or organisation.