Artificial intelligence tools have rapidly become part of education platforms, social media and entertainment services, making them a regular element of children’s and teenagers’ digital routines. Legal and data-protection frameworks have been slower to adapt, creating particular risks for these developing age groups.
How widespread is usage, and what are the international concerns?
- According to Pew Research, 64% of American teenagers use some kind of chatbot, and 30% rely on these systems daily.
- OpenAI’s own measurements indicate that young users mainly turn to the platform to process study material.
- The World Economic Forum’s Global Risks Report 2026 notes that online harms rank 30th on a two-year risk horizon, while AI-related harms rise to 5th place on a ten-year horizon.
These data points highlight that technological diffusion entails substantial exposure, with technical, social, psychological and legal implications.
Regulatory responses and international trends
Regulatory measures are tightening globally: Australia, the United Kingdom, Spain and Singapore, among others, are imposing greater obligations on platforms, with stronger age-verification and child-protection requirements. The issue was also discussed at the World Economic Forum’s 2026 annual meeting.
Market responses: what are developers doing?
Tech companies have responded by making safety settings default and introducing targeted restrictions. A notable market step: OpenAI launched ChatGPT for Teens on August 18, 2026, with key features including:
- age-based default classification that treats accounts of those under 18 as teen profiles, and a prohibition on access for children under 13;
- measures to reduce anthropomorphization, such as a switch to disable human-voiced replies;
- usage limits and quiet periods to encourage breaks;
- a Study Mode that focuses on guidance to understanding rather than providing ready-made answers;
- a parental notification protocol that can alert a parent in certain high-risk situations (for example, signs of self-harm or eating disorders) without granting automatic access to the child’s full conversations.
These solutions are developed by commercial actors and are not flawless: they must be used, measured, critiqued and iteratively improved.
Practical risks
Generative models introduce new types of vulnerability:
- lack of critical control: confidently phrased but incorrect outputs can be accepted as authoritative by young users;
- privacy exposure: free-form interactions can lead users to share sensitive personal information;
- emotional dependence: chatbots that imitate human support can foster psychological dependence;
- proliferation of deepfakes and manipulated visual content as generative image tools become widely available.
Role of education, families and teachers
Experts emphasize preparing children to understand what AI can and cannot do, how to approach its outputs critically, and when not to share sensitive data. This goes beyond limiting screen time: important skills include asking good questions, verifying answers and using AI deliberately to support learning and thinking.
Schools will have a crucial role, but teachers must first be trained to use and teach these tools confidently and competently.
Summary and conclusions
Rising AI use among young people requires legal, technical and educational responses. Lawmakers, technology companies and educational stakeholders need to act in coordination: regulatory frameworks should reflect young people’s actual digital habits, market solutions must be continuously evaluated and refined, and families and schools should actively teach responsible use.
Kiss Gergely — chairman of Attrecto Zrt. and board member of Signifer Csoport — as the article’s author recommends treating this challenge as a shared, ongoing learning process for technology companies, parents and society at large.
Cover image for illustration.
