Survey on generative AI use by children and parents

A survey by YouGov for the European think tank Interface found that children over 11 use generative AI tools almost as much as their parents, though parents underestimate how often their children use them. The survey, conducted in France, involved 1,138 parents and 970 adolescents aged 11 to 17, with parental consent.

Interface think tank's background

Interface, the think tank that commissioned the survey, describes itself as non-partisan and unaffiliated with any public or private institution. It has received funding from several organizations, including the Open Society Foundation, the William and Flora Hewlett Foundation, the Mercator Foundation, and the company Luminate.

Focus on generative AI use and perception

The survey aimed to understand how parents and children access and use generative AI, as well as how they perceive its impact. It found that most parents are familiar with general-purpose chatbots like ChatGPT and voice assistants, though fewer are familiar with AI companions like Character.AI or Nomi.

Parents underestimate children’s AI use

Parents significantly underestimate how often their children use generative AI. While 66% of parents believe their children use general-purpose AI, 78% of children say they do. The gap is smaller for other tools, but parents still underestimate weekly use by about 10 percentage points.

Parents see more risks than children

Parents are more cautious about the effects of generative AI on their children. Many believe it could hinder personal reflection or expose children to false information, while others see it as helpful for academic success. Children, however, are more positive, with nearly half saying AI saves them time on homework and fewer believing it prevents independent thinking.

Children want to use AI less often

Despite their frequent use of generative AI, 92% of children say they would like to spend less time on these tools. Many feel they use AI companions and social media-integrated AI too much, with nearly half expressing this concern about general-purpose AI and voice assistants.