New York bans AI in primary school, three months after Norway did

New York City is imposing a one-year moratorium on generative AI for pupils from 2-K through eighth grade, removing student-facing AI software and companion chatbots for around 600,000 children. Norway banned generative AI in its primary schools in June, and the EU AI Act regulates AI used on pupils rather than AI used by them.

New York City is barring generative AI for pupils from pre-school through eighth grade. The moratorium covers around 600,000 children and starts with the 2026-27 school year, ABC News reported.

Mayor Zohran Mamdani framed it as a refusal. “The tech industry wants us to believe that A.I.-powered early education is not only inevitable, but necessary“, he said. “We do not see it that way.”

The district is removing software rather than issuing guidance. All student-facing generative AI is being eliminated and companion chatbots are banned outright, the mayor’s office said.

High schools are exempt, but narrowly. Up to 50,000 pupils can join five pilot programmes, capped at five classes per school, with twice-yearly literacy modules for everyone else.

This is not the first such ban. Norway banned it in June for grades one to seven, with supervised use only up to 16.

That ban took effect at the end of August. Prime Minister Jonas Gahr Store said AI increases the risk of children skipping important steps in their education.

Europe therefore has the bans already. What it does not have is an AI rulebook that reaches the question.

The AI Act classifies education uses as high risk under Annex III. That means admission decisions, evaluation of learning outcomes, assessment of the appropriate level of education, and monitoring during tests.

Every one of those is AI applied to a pupil by an institution. The rulebook covers the system that admits a child and the one detecting prohibited behaviour in an exam hall, not the chatbot writing the essay.

Those obligations are also not in force yet. The Digital Omnibus pushed the Annex III deadline back to 2 December 2027.

One member state stopped waiting. Italy’s Law 132/2025 has required parental consent for under-14s to use AI since October last year.

The only EU-wide duty already binding is literacy. Article 4 has obliged deployers to ensure sufficient AI literacy since February 2025, which makes New York’s high school module a requirement Europe wrote first.

So the pattern repeats. Restrictions on children and AI keep arriving from education ministries and national parliaments, as they did when Italy set a worrying precedent in 2023, rather than from Brussels.