Months before New York City barred generative AI from classrooms, Norway had already implemented a nationwide ban, underscoring a significant gap in the European Union's AI regulatory framework. The Norwegian prohibition, effective from the end of August, applies to grades one through seven and allows only supervised use for students up to age 16.
Prime Minister Jonas Gahr Støre has voiced concerns that AI could allow children to bypass essential steps in learning to read, write, and perform basic mathematics—worries that have since resonated across the Atlantic. The move has been widely supported on social media platforms, including X, and has positioned Norway as a pioneer in restricting AI in educational settings.
EU's AI Act: A regulatory blind spot
While Norway is not an EU member, its decision highlights a critical omission in the bloc's AI Act. The legislation classifies education as "high-risk" under Annex III, but this designation only applies to systems used by institutions on pupils, such as admissions decisions, exam monitoring, and assessment of learning outcomes. It does not cover generative AI tools that students use themselves, such as chatbots for homework assistance.
Moreover, the compliance deadline for Annex III obligations has been pushed back to December 2027 due to the "Digital Omnibus" revision, leaving a prolonged period of regulatory uncertainty. However, one EU-wide rule is already binding: Article 4 of the AI Act, in force since February 2025, requires companies deploying AI to ensure "sufficient AI literacy"—a requirement that predates and closely resembles the literacy classes New York is only now introducing.
Italy's alternative approach
Italy has adopted a softer but still restrictive stance. Its Law 132/2025, effective since October last year, mandates parental consent for under-14s to use AI, rather than an outright ban. This represents a second, distinct model for regulating AI in classrooms within the EU, offering a contrast to Norway's hardline approach.
New York catches up
New York City has become the largest school district in the United States to follow Europe's lead. Mayor Zohran Mamdani and Schools Chancellor Kamar Samuels announced a one-year moratorium on student-facing generative AI for children from 2-K through eighth grade, affecting nearly 600,000 pupils—roughly two-thirds of the city's total public school enrolment.
The city will also disable AI functions in more than 38 previously approved classroom programmes that do not meet its new safety standards and will bar companion chatbots across all grade levels. Teachers will still be permitted to use AI for lesson planning, but not for grading or crisis management. High schools are only narrowly exempt: up to 50,000 students can join closely supervised pilot programmes, capped at five classes per school, while all other high schoolers will receive twice-yearly AI literacy lessons.
"The tech industry wants us to believe that AI-powered early education is not only inevitable, but necessary," Mamdani said. "We do not see it that way."
A patchwork response in the US
New York's moratorium follows a similar, if less formal, move by the Los Angeles Unified School District, the country's second-largest, which is temporarily blocking access to generative AI tools on student devices across all grades while it develops a permanent policy. Los Angeles also banned school-issued devices for its youngest pupils entirely in June as part of a broader screen-time policy.
At least 37 US states have issued official AI guidance for schools, though there is no shared definition of AI literacy or agreement on how it should be taught. New York's plan has drawn mixed reactions. Michael Mulgrew, president of the United Federation of Teachers, welcomed the new screen-time limits but said the AI policy "leaves many questions unanswered," including how the city's Education Department will ensure the software it buys has adequate safeguards.
Josh Golin, executive director of the children's advocacy group Fairplay, said a one-year review period was too short, arguing the burden should fall on AI companies to prove their products are safe, effective, and do not encourage cheating.
As Europe and the US grapple with the role of AI in education, Norway's early action and the EU's regulatory gaps offer a cautionary tale. The debate is far from settled, and the coming years will likely see further adjustments to policies on both sides of the Atlantic. For now, the focus remains on protecting foundational learning while navigating the rapid advancement of technology.
