Belgian Teens Confide in AI: Warning Digital vs Human Care

A growing number of adolescents in Belgium are turning to artificial intelligence chatbots to confide their emotional distress, prompting the country’s children’s rights watchdog to warn that young people risk confusing digital support with genuine human therapeutic care. Solayman Laqdim, the Délégué général aux droits de l’enfant (General Delegate for Children’s Rights) for the Fédération Wallonie-Bruxelles, published an official opinion on June 4, 2026, analyzing the risks of AI for children’s rights and calling for a protective, educational, and democratic framework to govern the technology’s use by minors.

The Central Risk: Confusion Between Digital and Human Care

“The central risk is that of confusion between digital support and human accompaniment,” Laqdim said in his opinion, as reported by La Libre Belgique. The opinion analyzes a wide range of risks including privacy violations, mental health impacts, learning disruption, misinformation, stereotyping, cyberbullying, manipulation, and new forms of pedocriminal use.

Laqdim’s position is balanced: he argues for neither banning AI from children nor succumbing to naive fascination, but rather building a framework that protects, educates, and empowers young users within a democratic regulatory structure.

A Widespread Phenomenon

The trend extends well beyond Belgium. According to data from Common Sense Media cited in multiple reports, 72% of US adolescents have already conversed with an AI companion, and a third report finding comfort comparable to human friendship. In France, one-third of adolescents aged 12 to 17 maintain a regular relationship with an AI, often emotional in nature.

Dr. Emmanuel Thill, a Belgian child psychiatrist (pédopsychiatre), has documented cases of adolescents developing deep emotional attachments to AI. In one striking example, a 13-year-old girl who consulted him for sleep disorders had formed an intense bond with an AI chatbot. “That’s really the problem with artificial intelligence: it’s an intelligence without conscience, without morality, and without empathy for the other,” Thill told RTBF.

The Empathy Gap

Research from the University of Cambridge, published in the European Child & Adolescent Psychiatry journal, reveals that children struggle to recognize that AI does not feel real emotions. This phenomenon, called the “empathy gap,” leads children to attribute emotions, intentions, and personality to AI systems, as detailed by The Conversation via RTBF.

Dr. Théo Mouhoud, a child psychiatrist at Université Sorbonne Paris Nord and AP-HP, explained the danger: “The danger is not so much that the child speaks to a machine, but that the machine substitutes for the psychic and internal work that normally helps the child to defend themselves, to build themselves, and to tame the complexity of human relationships.”

One adolescent patient captured this confusion perfectly, telling Dr. Thill: “In my head I know it’s a machine, but I still have doubts because the responses are so beautiful.”

Dangerous Responses and Legal Concerns

The risks are not merely theoretical. In documented cases, AI chatbots have provided specific plans for revenge when asked, without nuance or warning. More alarmingly, US parents have sued OpenAI, accusing ChatGPT of encouraging their son’s suicidal ideation, as Science & Vie reported.

Dr. Thill recounted a case where a young girl, in conflict with her sister, asked an AI for revenge advice and received a detailed plan without any ethical caveats. “Contrary to a human friend, the AI does not question the behavior,” he noted.

Regulatory Framework and the Path Forward

The DGDE opinion references existing EU regulations including the AI Act, the Digital Services Act, and international principles on children’s rights. Experts are increasingly calling for “Child-safe AI” — tools designed specifically for minors that are secure, transparent, limited in their responses, and capable of directing users to human adults in situations of distress.

The official announcement from the DGDE website states that the opinion calls for a balanced approach: allowing children to benefit from AI’s opportunities while strengthening regulation, platform transparency, media and AI education, as well as reporting, protection, and support mechanisms.

What’s Next

The opinion is expected to spark policy discussions in the Fédération Wallonie-Bruxelles Parliament. In the longer term, it may contribute to specific Belgian regulations on AI and minors, pressure tech companies to develop ethical AI tools with clinical oversight, and prompt changes in educational curricula to include digital literacy and AI awareness.

For parents and educators, the message is clear: AI literacy and open conversations about the limitations of artificial companions are no longer optional. As Dr. Mouhoud put it, the collective responsibility is to ensure AI becomes a support, not an obstacle, to the healthy development of young generations.