About the risks posed by children’s friendship with AI and how to protect them from it, speaks Jurgita Smiltė Jasiulionė, psychologist at “Vaikų linija”, and Julija Markeliūnė, Head of Digital Inclusion at “Telia”.
Millions of children every day not only do their homework with AI but also talk about their daily lives. The UNICEF Innocenti office conducted a study at the beginning of 2026, which showed that in some countries, up to 50% of surveyed children already use AI, and approximately 1 out of 10 children turn to a chatbot for personal problems or concerns, seeking advice and emotional support.
The study also revealed that children are more than three times more active AI users than their parents or guardians, thus a digital competence gap remains in families. This time, the study examined data collected from 10 different countries around the world, such as Armenia, Brazil, Mexico, Serbia, Montenegro, etc.
Start trusting AI more than loved ones
Experts point out that despite AI’s significant benefits for learning, creativity development, and information accessibility, its use is also associated with potential emotional harm.
“AI is easily accessible, and children tend to use it in a variety of situations, but due to developmental characteristics, it is difficult for children to critically evaluate the information it provides. Frequent use of AI for learning tasks weakens a child’s independent thinking and problem-solving skills. And its use for emotional support sometimes leads to the creation of an emotional bond with chatbots,” says Jurgita Smiltė Jasiulionė, psychologist at “Vaikų linija”.
According to her, by engaging in close communication with chatbots, children may trust their answers more than the advice of close people. This disrupts healthy communication and the development of interpersonal relationships.
“It is important to explain that AI is not a real friend or psychologist, and that they should critically evaluate the information and advice it provides. It is essential to encourage children to seek emotional support, advice, and help primarily from family members, teachers, or other trusted adults. Open communication and trust allow children to more easily recognize threats and to question the answers they receive when talking to AI,” explains the psychologist.
Read more Drunk drivers on Polish roads: sentenced in an expedited procedure
Reveal too much personal data
According to Julija Markeliūnė, Head of Digital Inclusion at “Telia”, concern should also be raised because consultations with chatbots do not always provide accurate and correct information. Children are not always able to critically evaluate the reliability of the answers provided, and it is also difficult for them to distinguish manipulations – for example, so-called “deepfake” images.
“When communicating with chatbots, children reveal a lot of personal information and do not consider how it might be used. For example, one of the new dangers is so-called “undressing” or “nudification,” where AI tools are used to digitally remove or alter clothing in existing photos, thereby creating fake images of nude people. This poses a new threat of child sexual exploitation, blackmail, and bullying,” J. Markeliūnė shared her insights.
The interviewee also draws attention to the reasons why children like to communicate with chatbots so much. After all, AI always agrees with us, practically never contradicts, constantly praises – even adults get hooked on this.
“Children, especially teenagers, often feel misunderstood at certain stages of development and believe that the whole world is against them. At such moments, AI can become their 100% supporter, but this can further increase the disconnect between the perception of reality and what AI provides,” emphasizes J. Markeliūnė.
The expert advises parents to review and take care of their child’s digital environment: use parental control tools, set age-appropriate privacy settings, and regularly inquire about what apps or AI tools their child is using.
“Of course, it is still very important not only to rely on cybersecurity measures but also to teach children not to share personal information, photos, addresses, phone numbers, or other sensitive data in conversations with AI,” emphasizes J. Markeliūnė.
Read more Violin celebrity Akiko Suwanai, captivating with exceptional virtuosity, will perform in Lithuania
