A quarter of UK pre-teens are now using GenAI

While adults are busy worrying about AI stealing jobs, kids have quietly turned it into a creative toy, a study buddy, and in some cases, a digital confidant.

A new report from The Alan Turing Institute surveyed 780 children aged 8 to 12 alongside their parents, plus 1,001 primary and secondary school teachers across the UK. The findings offer one of the clearest looks yet at how generative AI is actually taking root in family life.

Homework, doodles, and digital companionship

According to the report, 55% of UK households now use generative AI tools. Among children aged 8 to 12, nearly a quarter report having used them directly, with 72% of those users logging on at least once a month.

Adoption climbs steadily with age, rising from 14% among 8-year-olds to 26% among 12-year-olds, with a slight plateau at age 11. Girls reported slightly higher usage than boys, at 24% compared to 19%.

When it comes to specific platforms, ChatGPT leads the pack with 58% of young users, followed by Google Gemini at 33% and Snapchat's My AI at 27%.

What kids do with AI shifts as they grow. Eight-year-olds primarily use it for entertainment, nine-year-olds use it to look up information, ten-year-olds generate funny pictures, eleven-year-olds mix play with research, and by age twelve, schoolwork and homework assistance become the number one use case.

The data also revealed a striking trend among children with additional learning needs. In this group, ChatGPT usage hit 78%, compared to 53% among their peers.

More importantly, children with learning needs rely on AI far more for connection. Around 39% turn to AI for personal advice, compared to 16% of other children. Another 37% use it for companionship, and 30% use it to play with friends.

The emerging household divide

AI adoption is not spreading equally across the board. The research found sharp divides across geography and socioeconomic lines.

In England, 57% of parents reported household AI use, while numbers fell to 50% in Wales, 41.2% in Northern Ireland, and 40% in Scotland.

Income and social grade created an even wider gap. While 61% of households in social grades A, B, and C1 reported using generative AI, that figure dropped to 44% for grades C2, D, and E.

That disparity filters down directly to kids. Among children who had even heard the term generative AI, 70% came from higher socioeconomic grades, versus only 30% from lower grades. Awareness was also notably lower among state school students compared to those attending private schools.

Parental attitudes play an outsized role as well. Among the children most enthusiastic about AI, 93% have parents who view their AI usage positively. In contrast, children who do not use the technology are far more likely to find it confusing or intimidating.

Why it matters

The survey challenges the common narrative that children only use AI to cheat on essays or consume automated content.

Instead, young users treat AI as an interactive canvas and an encyclopedic playground. Perhaps most compelling is its role as an empathetic buffer for neurodivergent kids or those with extra learning needs, offering a judgment-free space to ask sensitive questions and practice communication.

At the same time, we think the report exposes a dangerous blind spot in education policy. The divide between private and state schools, and across household income brackets, suggests that early AI literacy is already becoming a luxury good.

If schools and governments do not step in to provide structured access and guidance, generative AI could rapidly widen the developmental gap between children who are coached into tech fluency and those left behind.