Children adopt AI three times faster than adults: the challenge of governing digital childhood

What happened?

UNICEF has published an analysis based on data from 10 countries — Australia, Brazil, Chile, the United States, India, Indonesia, Japan, Kenya, the United Kingdom, and South Africa — revealing that approximately 20 million children have already used artificial intelligence tools. The adoption rate among minors is three times higher than among adults, leading the organization to describe this situation as a 'global experiment' in which an entire generation grows up immersed in AI without adequate safeguards. The report, titled 'Children and AI: A Global Experiment,' was presented in July 2026 and is based on surveys of over 10,000 households, as well as usage data from platforms such as ChatGPT, Bard, and other generative tools. UNICEF warns that, unlike previous technologies, generative AI is particularly attractive to children due to its interactivity and personalization capabilities, accelerating its adoption.

Why is it important?

The speed of AI adoption by children far exceeds the capacity of governments and institutions to establish effective governance frameworks. Unlike previous technologies, such as the Internet or smartphones, AI poses specific risks: algorithmic biases, exposure to inappropriate content, behavioral manipulation, and mass data collection without informed consent. Children, being in the midst of cognitive and emotional development, are particularly vulnerable to these effects. For example, a 2025 Stanford University study showed that chatbots can influence minors' moral decisions, while the U.S. Federal Trade Commission has documented cases of data collection from children without parental permission through voice assistants. The OECD, in its 2024 report, already noted that only 12% of member countries had specific policies for AI and childhood. This regulatory gap contrasts with rapid penetration: according to UNICEF, 40% of children aged 12-17 in developed countries already use AI weekly.

Consequences and challenges

The consequences of this early adoption can be profound. On one hand, AI offers unprecedented educational and creative opportunities: personalized tutors, translation tools, homework assistants, and artistic creation platforms. Countries like Finland have integrated AI into the school curriculum since 2024, with promising results in critical thinking skills. However, the lack of specific regulation for minors exposes children to risks that could affect their privacy, mental health, and social development. A 2025 Common Sense Media study found that 30% of children using chatbots have unknowingly shared sensitive personal information. Additionally, algorithmic biases can reinforce gender and racial stereotypes, as evidenced in a University of Cambridge experiment where a generative AI produced biased images when asked to draw 'a smart boy' or 'a pretty girl.' UNICEF urges governments to update child protection laws to include AI, as well as technology companies to implement ethical designs from the ground up. The challenge is urgent: while adults debate, children are already using these tools. The European Union, with its AI Act, is a pioneer in classifying AI systems targeting children as high-risk, but effective implementation is not expected until 2028.

What readers should know

The UNICEF report not only warns about the speed of adoption but also highlights the lack of data on long-term impact. Parents and educators should inform themselves about the tools children use, encourage critical use, and demand transparency from platforms. The governance of children's AI is not a future issue: it is a present necessity. Compared to the arrival of the Internet in the 1990s, when governments took a decade to establish online child protection laws, AI is advancing much faster: in just three years, usage among children has tripled. Organizations like UNESCO have already published guidelines for AI in education, but their adoption is voluntary. Responsibility also falls on companies: OpenAI, for example, has introduced age restrictions for ChatGPT (13 years with parental permission), but UNICEF notes that these measures are easy to bypass. The call is for coordinated action: regulation, education, and ethical design so that AI is not an experiment but a safe tool for future generations.