BERLIN, July 9 (Xinhua) -- As generative Artificial Intelligence (AI) reaches children through both the classroom and the internet, governments and schools are attempting to simultaneously protect young users and prepare them for a technology that will become increasingly necessary in the future.
The debate over children's use of AI has intensified recently as the United Nations (UN) has called for stricter child-safety rules for AI systems, while several European countries have moved to define when and how students should be allowed to use the technology in schools.
FROM GLOBAL WARNINGS TO CLASSROOM RULES
UN Secretary-General Antonio Guterres on Monday warned that AI is reaching children faster than safeguards can be put in place to protect them.
Speaking at the opening of the Global Dialogue on AI Governance in Geneva, Guterres underlined that products other than AI face strict safety checks before being marketed to children.
"We test every toy before it reaches a child ... Yet AI has reached our children -- their learning, their friendships, their most private questions -- before anyone asked what it would do to them," he said.
The risks of AI are no longer theoretical, he stressed, warning that children are already being deceived by machines posing as friends, steered toward self-harm and harmed by images of abuse that can be created at the touch of a button.
"No child should be a guinea pig for unregulated AI," he stressed.
That concern is already shaping national policy. In Norway, the government announced new guidelines in June that would strictly limit the use of generative AI in schools, especially among younger children.
Under the plan, pupils in grades one to seven will generally not be given access to AI from the autumn school term. Students in grades eight to 10, meanwhile, may use AI gradually and cautiously, provided that teachers have first acquired sufficient competence before students receive training and access. Upper secondary students are expected to learn how to use AI appropriately so they are prepared for further education and work.
France has already moved in a similar direction in 2025, when it issued an official framework, saying that students may use generative AI in class only from France's 4e onward, roughly equivalent to eighth grade, and only when its use is limited, supervised, explained and guided by the teacher.
WHAT IS BEHIND CALLS FOR AI RESTRICTION
For schools, the concern is not only what AI may expose children to, but also what it may take away from the learning process.
Norway said it has framed the AI restrictions as part of a wider effort to restore basic learning in schools after years of heavy reliance on screens and digital materials. The government said learning outcomes and basic skills in Norwegian schools were declining, and called for less uncritical use of AI and screens, more physical books, and stronger reading, writing and numeracy skills.
"We have a major challenge that we are now addressing: many children and young people are unable to write, read and do arithmetic well enough," Norwegian Prime Minister Jonas Gahr Store said in a separate speech. "We know that uncritical use of AI causes students to skip important steps in learning."
This also points to a more familiar problem for schools: whether teachers can still tell what work is a student's own. According to the Organization for Economic Cooperation and Development's latest report on digital education outlook, 72 percent of lower secondary teachers believe AI can harm academic integrity by allowing students to pass off AI-generated work as their own.
Stefan Duell, president of the German Teachers' Association, echoed this concern. "Anything that happens outside the classroom could theoretically have been completed with or by AI," which made authorship difficult for teachers to verify, he said.
For that reason, Duell said schools should require handwritten work for homework and exams in lower grades to help prevent cheating, and teachers need to pay closer attention to the process behind students' answers.
DRAWING GUARDRAILS FOR AI IN SCHOOLS
As AI digs itself ever-more deeply into daily life, keeping students entirely away from it is becoming increasingly unrealistic.
Annekatrin Bock, a professor of media research and the digitalization of education at the University of Vechta in Germany, has argued that AI should not simply be kept out of classrooms.
"The topic of AI belongs in schools," she said, while stressing that it must be introduced in an age-appropriate way and embedded in teaching from a pedagogical and didactic perspective.
Bock said bans or a purely protective approach would do little good, because students would then explore AI privately and without guidance. Even in primary school, she said, children could use AI to generate images and then write character descriptions based on them.
However, supervised use in classrooms addresses only part of the problem. In Geneva, Guterres called for an AI Child Safety Pledge with rules for any system accessible by children.
Such a pledge, he said, should require companies to prove that AI systems accessible to children are safe through child-specific testing and independent oversight. It should also impose zero tolerance for sexual abuse material, requiring companies to prevent AI from generating sexual images of children and to detect, report and remove such content.
Guterres said AI systems must not leave children in crisis alone. When a child shows signs of distress, he said, the system should stop and connect the child to real human support.
"When a child is harmed, the answer must never be 'the algorithm did it,'" he added. ■
