Children and AI: Age-specific risks and practical guidance for families

By Ahmet Taş | Wise News Press

IZMIR, TURKEY — Artificial intelligence can support children’s learning and creativity, but unsupervised or developmentally inappropriate use may expose them to misinformation, privacy concerns, emotional dependency and reduced real-world interaction.

Dr. Sema Bayçın, a psychiatry specialist at Batıgöz Balçova Surgical Medical Center in Izmir, said families should not approach AI solely through prohibition. Instead, parents should guide children, establish age-appropriate limits and ensure that digital tools do not replace play, family communication or face-to-face relationships. Batıgöz lists Bayçın as a psychiatrist at its Balçova center.

A 2026 review published by the American Academy of Pediatrics said generative AI is becoming increasingly embedded in children’s learning, play and everyday life. The review identified potential benefits such as personalized education and creative expression, alongside risks including false information, privacy threats and children perceiving AI as a friend or caregiver.

Real-world experiences should come first in early childhood

During the first years of life, children learn primarily through physical exploration, play, movement and responsive interaction with parents and caregivers.

When screens occupy too much of a young child’s day, they may displace time that would otherwise be spent talking, playing, sleeping or interacting with other people. The effect can depend not only on how long a screen is used, but also on what the child is doing, whether an adult is present and what activity the screen replaces.

World Health Organization guidelines recommend no sedentary screen time for infants and 1-year-olds, while screen time for children aged 2 and those aged 3–4 should be limited to no more than one hour a day, with less considered preferable. WHO emphasizes physical activity, adequate sleep and interactive activities such as reading and storytelling.

A cohort study involving 7,097 mother-child pairs found an association between longer screen exposure at age 1 and delays in communication and problem-solving at ages 2 and 4. The researchers stressed that the study showed an association rather than proving that screen use directly caused the delays, and they could not separate educational screen use from other forms of screen exposure.

“In early childhood, children’s most important source of learning is real-life experience,” Bayçın said. “Playing, communicating with peers, spending time with family and exploring the environment are fundamental to social, emotional and cognitive development.”

Preschool children may mistake AI for a living being

Children between roughly 2 and 6 years old may not always clearly distinguish between a machine that imitates conversation and a person who thinks, feels and understands.

Voice assistants, talking toys and conversational AI systems can respond in natural language, remember previous exchanges and appear emotionally responsive. Bayçın said some young children may therefore interpret such systems as living beings or social companions.

This does not mean children must be completely separated from AI-based educational tools. UNICEF advises parents to begin discussing AI at an early age because children may already encounter it through phones, toys, schools and household devices. Adults can explain in simple language that an AI system does not have human feelings and that its answers may be incomplete or wrong.

For preschool children, shared use is particularly important. A parent can sit beside the child, ask what the system is doing, correct misleading information and connect the digital activity to drawing, storytelling, building or outdoor play.

Young children should also be taught not to give a chatbot their full name, home address, school, photographs or family information. UNICEF’s child-centered AI guidance places safety, privacy, transparency and healthy development among the main requirements for systems used by or affecting children.

Primary school children need help checking AI answers

Children in primary school may use generative AI to explain homework, create stories, practise languages or explore unfamiliar subjects.

Used carefully, such tools can provide examples, suggest questions and help children approach a difficult topic from another angle. The risk arises when the system completes the entire task or when the child assumes that every fluent answer is accurate.

Generative AI can produce fabricated facts, nonexistent references and misleading explanations. The American Academy of Pediatrics review therefore recommends building AI literacy and critical thinking rather than treating AI output as automatically reliable.

Parents and teachers can encourage children to ask several questions before using an answer:

Where did this information come from? Can it be confirmed through a textbook or reliable website? Does another source say the same thing? Which part of the final work reflects the child’s own thinking?

Bayçın warned that repeatedly receiving ready-made answers could reduce opportunities for problem-solving, independent writing and creative experimentation. A safer approach is to use AI for brainstorming, explanations or feedback while requiring the child to research, decide and produce the final work.

AI companions may create additional risks during adolescence

Adolescence is a period in which identity, peer approval, emotional intimacy and independence become especially important.

AI companions can feel easier than human relationships because they are continuously available, rarely disagree and can be designed to offer reassuring responses. For some young people, that may encourage prolonged engagement or emotional dependence.

UNICEF said in a June 2026 policy brief that children and young people increasingly use AI chatbots not only for information and creativity, but also for advice, emotional support and relationships. The organization identified conversational and companion-style AI as presenting distinct and potentially heightened risks for children.

These risks do not mean that every adolescent using a chatbot will become socially isolated. Evidence about the long-term effect of AI on children’s social, emotional and cognitive development remains limited, according to UNICEF.

Parents should nevertheless pay attention when a young person begins withdrawing from friends, hiding extensive chatbot use, losing sleep or treating an AI character as the only relationship in which they feel understood.

Adolescents should also know that personal conversations may be collected, stored or used by the service provider. Health concerns, family conflicts, intimate images and identifying information should not be shared with AI systems as though they were confidential doctors or therapists.

Deepfakes and AI-generated content can intensify cyberbullying

AI image, audio and video tools allow young users to create realistic content rapidly. These tools can support art, media literacy and school projects, but they can also be used to impersonate classmates, alter photographs or produce humiliating and non-consensual material.

UNICEF’s updated guidance highlights risks involving harmful deepfakes, exploitative material and AI-generated intimate images. It recommends preventive safeguards, accountability and stronger protection of children’s data and dignity.

Families and schools should make clear that creating or sharing a manipulated image of another person without permission can cause serious emotional harm and may also have disciplinary or legal consequences.

Children who are targeted should be encouraged to preserve evidence, block the account, report the content to the platform and inform a trusted adult. Threats, sexualized images, blackmail or immediate safety concerns require prompt professional and institutional support.

Parents should guide rather than simply ban AI

Bayçın said AI should be treated neither as entirely harmful nor as automatically beneficial.

Age-appropriate and supervised use can support language learning, coding, accessibility, research and creative work. The central question is whether AI assists the child’s thinking or replaces it.

Families can establish clear rules about when AI may be used, which applications are permitted and what information must never be entered. Younger children should use AI with an adult, while older children should learn to verify claims and disclose when AI has contributed to schoolwork.

Technology should not interfere with sleep, physical activity, meals, family conversation, friendships or responsibilities. Parents can also model healthy behaviour by limiting their own device use during shared time.

“AI and digital tools should not replace play, real social relationships, family communication or productive learning experiences,” Bayçın said.

UNICEF’s guidance similarly calls for AI systems that support children’s best interests, development and wellbeing while protecting privacy, safety, fairness and inclusion.

A child or adolescent who becomes unable to reduce AI use, withdraws from real relationships, develops significant distress or relies on a chatbot for serious mental-health advice may benefit from assessment by a qualified child and adolescent mental-health professional.

WiseNewsPress.com