New York prohibits the use of AI in primary schools, three months after Norway implemented a similar ban.

New York prohibits the use of AI in primary schools, three months after Norway implemented a similar ban.

      New York City is implementing a one-year moratorium on generative AI for students from pre-kindergarten to eighth grade, eliminating student-accessible AI tools and companion chatbots for approximately 600,000 children. In June, Norway also prohibited generative AI in primary schools, while the EU AI Act regulates AI applications for students rather than their usage.

      According to ABC News, New York City's ban will affect students from pre-school through eighth grade, commencing with the 2026-27 school year. Mayor Zohran Mamdani described this decision as a rejection of the tech industry's narrative. “The tech industry wants us to believe that A.I.-powered early education is not only inevitable, but necessary,” he stated. “We do not see it that way.”

      Rather than providing guidance, the district is removing software entirely. All generative AI that students can access will be eliminated, and companion chatbots will be completely forbidden, according to the mayor's office.

      High schools are exempt from this ban, albeit narrowly. Up to 50,000 students may participate in five pilot programs, limited to five classes per school, while literacy modules will be required twice a year for all other students.

      This is not the first prohibition of its kind; Norway implemented a similar ban for grades one to seven in June, allowing supervised use only for students up to age 16. This ban went into effect at the end of August. Prime Minister Jonas Gahr Store stated that AI could lead to children bypassing essential aspects of their education.

      Europe has already seen such prohibitions, but it lacks a comprehensive AI regulatory framework addressing this issue. The AI Act categorizes educational applications as high risk under Annex III, covering areas like admission processes, evaluation of learning outcomes, assessment of suitable education levels, and monitoring during exams. These obligations pertain to AI's influence on students as administered by educational institutions, rather than AI tools like chatbots that assist with writing essays.

      Furthermore, these obligations are not yet in effect, as the Digital Omnibus has postponed the Annex III deadline to 2 December 2027. One member state, however, decided not to wait; Italy’s Law 132/2025 has mandated parental consent for the use of AI by children under 14 since October of last year.

      The only binding EU-wide requirement currently in place is related to literacy. Article 4 has required those deploying AI to ensure adequate AI literacy since February 2025, which positions New York's high school program as a necessity that Europe had envisioned first.

      Thus, the trend continues, with education ministries and national parliaments imposing restrictions on children and AI, similar to Italy’s concerning precedent set in 2023, rather than resulting from initiatives in Brussels.

Other articles

The gap in AI skills is widening as schools anticipate flawless solutions. The gap in AI skills is widening as schools anticipate flawless solutions. According to Gallup, teachers who utilize AI on a weekly basis save an average of 5.9 hours each week. The World Economic Forum states that by 2030, 39% of skills could become obsolete. Kirk Drake contends that instead of prohibiting AI, schools should focus on educating teachers to use it responsibly. The Justice Department supports OpenAI in the copyright dispute with publishers. The Justice Department supports OpenAI in the copyright dispute with publishers. The DOJ informed a court in New York that AI training constitutes fair use and pertains to national security. Recital 106 of the AI Act was specifically created to prevent such situations. Google has launched Gemini 3.8 Flash alongside a cybersecurity version restricted to governmental use. Google has launched Gemini 3.8 Flash alongside a cybersecurity version restricted to governmental use. Google has launched Gemini 3.8 Flash and Flash Cyber. Each general-purpose model available in the EU market is subject to Article 53 obligations, while the cyber model is restricted. What occurs to trust when focus takes precedence over honesty? What occurs to trust when focus takes precedence over honesty? Booker Loud, the founder of Pin-Social, asserts that the attention economy of social media is undermining public trust. With only 37% of people worldwide trusting news and 73% of Americans unable to tell real information from fake online, he advocates for digital environments that prioritize connection instead of outrage. A US court has ruled that Google can maintain its ad exchange. The Commission stated that a sale is the only viable solution. A US court has ruled that Google can maintain its ad exchange. The Commission stated that a sale is the only viable solution. Google will not be required to sell its ad exchange in the US. Officials in Brussels state that only a partial divestment of its services would address the same behavior. Acer's latest Vero 16 demonstrates that laptops designed for easy repair can still offer a satisfying experience. Acer's latest Vero 16 demonstrates that laptops designed for easy repair can still offer a satisfying experience. Acer's revamped Vero 16 features a chassis made from recycled aluminum that is easy to repair, paired with top-tier Intel processors, demonstrating that eco-friendly laptops can deliver high performance.

New York prohibits the use of AI in primary schools, three months after Norway implemented a similar ban.

New York City is prohibiting generative AI for students up to the eighth grade. Europe has already implemented two similar bans, along with an AI Act that does not cover the students utilizing it.