One in Four Young Harassment Victims in France Turned to AI for Support, e-Enfance Barometer Finds

A quarter of French children and teenagers who have been harassed say they turned to an artificial intelligence tool for support, according to the 2026 barometer published on October 29 by the child-protection charity Association e-Enfance with the bank Caisse d’Epargne. The survey, carried out by the polling firm George(s) from September 18 to 28, questioned 1,612 pairs of parents and children aged 6 to 18. RFI, which reported the findings, described AI as becoming a confidant for young people.

According to the study, 32% of young people say they have experienced harassment or cyberharassment, with 29% reporting harassment and 12% cyberharassment. Among the victims who used an AI tool for support, 83% said it helped them. AI is also part of everyday school life: 66% of middle and high school pupils say they use it for schoolwork, up from 57% in 2025, and 21% say they use it as a confidant.

Artificial intelligence also features in the harm itself. The barometer found that 31% of cyberharassment victims had been targeted with humiliating content or deepfakes generated by AI. Seven in ten victims told their parents about what happened, 64% said psychological support would have helped them, and 10% said they had thought about suicide.

What it means in practice

For parents and schools, the figures suggest chatbots are already filling a gap left by adults and support services. That raises questions about the advice these tools give a distressed child and whether they steer young users towards a trained person. Concern about how AI shapes young people’s prospects is also coming from other directions, as when a Banque de France study found permanent hiring of young people falling in AI-exposed jobs.

The deepfake findings add pressure on platforms to detect and remove AI-generated humiliation quickly, and on schools to treat such images as harassment rather than pranks. Public debate about who sets these rules is growing, with civic groups that planned a Paris rally calling for a citizens’ convention on AI.

What to watch next is whether the education ministry and the main platforms respond to the deepfake figures, whether AI companies add safeguards for minors who use chatbots for emotional support, and whether next year’s barometer shows more pupils confiding in AI or in adults.