French parents underestimate children’s AI use, survey finds

A new survey has found that 40% of children aged 11 to 12 in France use general-purpose artificial intelligence tools such as ChatGPT every week, while only 26% of their parents realise they do so.

The findings highlight a growing gap between children’s use of AI and their parents’ understanding of the technology, with 42% of parents saying they do not feel supported at all in helping their children use AI responsibly.

The survey was commissioned by Berlin-based non-profit Interface and conducted by YouGov between 24 and 30 June 2026. It included 1,138 parents and 970 children aged 11 to 17 in France.

Children are using AI from an early age.

The survey shows that 56% of children aged 11 to 12 use some type of generative AI weekly. In addition to the general-purpose chatbots, 33% use AI voice assistants, 23% use AI built into social media and 16% use AI companions.

Their uses vary, from homework and research help to conversations and companionship. In addition, there are AI functions incorporated in social media and games which allow children to access AI without even actively seeking out a chatbot.

The difference between the usage reported by children and estimated by parents remains consistent throughout all age groups. For 13- to 15-year-old children, 62% report regular usage of generative AI, whereas 52% of parents think their children use AI. For 16- to 17-year-old teenagers, the numbers were 70% and 61%, respectively.

Interface senior policy researcher Jessica Galissaire also highlighted concerns about the amount of time children spend using these services. According to the survey, 92% of generative AI users said they spent more time on the technology than they wanted to.

AI companions and AI features integrated into social media recorded particularly high figures, at 59% and 57%, respectively, among users who said they spent longer on them than intended.

The findings suggest many parents are struggling to keep pace with their children’s use of artificial intelligence. While 32% felt sufficiently supported in helping their children navigate AI, 26% said they received some support but not enough.

Galissaire said parents with limited digital knowledge or less time to monitor their children’s online activities could find it particularly difficult to discuss AI use and implement parental controls.

When asked who should help children understand and navigate AI, 51% of parents identified teachers and schools as responsible. Meanwhile, 41% pointed to technology platforms, while 30% said public authorities should play a role.

The results suggest parents expect responsibility to be shared across educational institutions, technology companies and governments rather than placed solely on families.

In September 2026, the European Commission made known its proposals for the EU Kids Act, which included the intention of strengthening platform liability and parental control mechanisms, as well as addressing any design flaws that may be harmful.

But Galissaire noted that relying only on age restrictions and parental control tools was not enough to keep children safe online. She preferred that there would be mandatory safe-by-design tools on the part of the platforms.