A Chilean engineer calls for copying the school AI moratorium New York applies to 600,000 children

By Zendoric · September 13, 2026. A column published in the Chilean outlet Alerta Noticias Temuco raises a concrete question: should Chile apply a school moratorium on generative artificial intelligence (AI)? The text—whose authorship is not specified in the source—proposes establishing a temporary pause in the country on the direct use of these tools in primary education, so their effects can be studied before they are normalized in the classroom.

The precedent usually invoked in this debate—and one we have followed closely at Zendoric—is that of school moratoriums on the direct use of generative AI in public schools; the available source does not make it possible to verify which specific measure the column cites, nor its duration, its scope or the number of students affected. The relevant nuance of this type of measure is that it does not ban AI across the education system as a whole—teachers are usually exempt and may use it as a support tool—but rather blocks younger children's direct access to chatbots and generative assistants while the impact is assessed.

The column backs the proposal with the recommendation UNESCO (United Nations Educational, Scientific and Cultural Organization) issued back in 2023: setting 13 as the minimum age for independent use of generative AI in educational settings, along with data protection, specific teacher training and ongoing pedagogical evaluation. It also cites, according to the text itself, a warning from Pope Leo XIV about the risk of confusing artificial intelligence with human intelligence, in line with the author's underlying argument: cognitive effort—getting things wrong, trying again, building an idea of one's own—is a constitutive part of learning and of children's socio-emotional development, and delegating it to a machine too early can erode skills that are still forming.

Here it is worth separating two things that are easily conflated: AI as a tool for an adult who already knows how to think for themselves, and AI as a substitute for the process by which a child learns to think. On the first, the evidence points to real gains in productivity and personalized learning; on the second, the column's argument is plausible even without conclusive data yet: a developing brain does not benefit from a shortcut in the same way as one already formed. That is why the serious debate is no longer "ban or allow," but at what age, with what pedagogical design and with what safeguards the tool is introduced.

That a Chilean opinion column leans on precedents arriving from abroad says something about how AI governance in education is spreading: not through binding international treaties, but through precedents that travel from city to city and from column to column, far faster than the ability of education systems like Chile's to fund the teacher training or the data protection that UNESCO itself demands as a condition, not an accessory. Copying the headline—«a pause on AI in the classroom»—is easy; copying the supporting infrastructure that sustains it, far less so, and that is where this kind of proposal tends to fall short.

Zendoric's underlying thesis does not change because of this: over the long term, a well-designed AI introduced into education with judgment can personalize learning and free teachers from administrative burden so they can devote that time to what an algorithm does not replace—mentoring, judgment, the human relationship with the student. But that horizon is not reached by skipping childhood: if cognitive effort is, as this column suggests, the material from which the capacity to think is built, delaying direct access in the early years is not a brake on progress, it is the condition for that progress to reach students capable of making the most of it with judgment of their own.