New York has just opened a debate that Chile should not evade. Its mayor, Zohran Mamdani, established a one-year moratorium on the direct use of generative artificial intelligence for nearly 600,000 students in public schools, from prekindergarten through eighth grade.
This is not about banning artificial intelligence, but about studying its effects before incorporating it massively at a decisive stage of human development: childhood.
Generative artificial intelligence is capable of creating content based on the instructions it receives: it can write, answer questions, solve problems, or create images, videos, and music. Precisely because it can do so many things that previously required human intellectual effort, we must ask ourselves what happens when we put that capability in the hands of children who are still developing the skills that the machine can replace.
The fundamental question is not technological. It is what we understand education to be.
School does not exist merely to transmit information. Education aims at the comprehensive development of girls and boys: teaching them to think, create, ask questions, make mistakes, persevere, relate to others, and solve problems.
Cognitive effort is a fundamental part of learning and socioemotional development. Searching for an answer and not finding it, making a mistake, becoming frustrated, trying another path, trying again, and finally understanding are educational experiences. The process matters more than the result.
An AI that instantly delivers an extraordinarily elaborate answer can be formidable for someone who has already developed those abilities. The risk arises when it replaces precisely the effort through which a child must acquire them.
The same applies to living together with others. We learn by talking, listening, disagreeing, cooperating, resolving conflicts, and building bonds. Empathy cannot be downloaded from the internet. It develops through relating to other people.
That is why the recommendation UNESCO has upheld since 2023 is pertinent: it proposes 13 as the minimum age for independent use of generative AI and calls for data protection, teacher training, and pedagogical evaluation. At a time when multilateral organizations are attacked by nationalists and right-wing sectors, it is worth reaffirming it. Protecting children’s development is not an ideological position.
Pope Leo XIV also warns that we must not confuse artificial intelligence with human intelligence: a machine processes information, but it does not have experiences or mature through relationships. Technological progress can end up being dehumanizing if we increase the power of machines while weakening human bonds.
When a laboratory develops a drug, demanding research and evidence regarding its safety does not mean being against innovation. The history of thalidomide demonstrates why such safeguards exist.
We do not know whether artificial intelligence will be the new thalidomide. The lesson is clear: when an innovation can profoundly affect human development, we should not wait to confirm the harm before investigating its effects and establishing precautionary limits.
Economic freedom and innovation are fundamental to progress, but they are not absolute values. The well-being and comprehensive development of girls, boys, and young people take precedence over the economic freedom to introduce a technology without sufficiently knowing its consequences.
Chile should apply the same precautionary principle: establish a temporary moratorium on the direct use of generative AI by children in basic education and use that period to study and evaluate its effects. Waiting for long-term studies while the technology is being incorporated massively means arriving too late.
The genie of artificial intelligence is already out of the bottle and will not go back in. These and future generations will have to learn to use it.
But there is an order we cannot reverse.
First we must form human beings recognized in their dignity, capable of thinking, creating, asking questions, making mistakes, persevering, relating to others, and solving problems. Then we must teach them to use artificial intelligence to enhance those abilities, not replace them.
Artificial intelligence will help us build a better education if we are capable of putting it also at the service of this purpose: that it contribute to developing fully human beings and that it itself help us avoid, by depending on its answers, ending up behaving like machines.
