New York City halts generative AI tools for 600,000 elementary and middle school students

Scope of the classroom moratorium

New York City Mayor Zohran Mamdani announced on Wednesday that public schools will impose a one-year moratorium on student-facing generative artificial intelligence starting with the 2026–2027 school year. The policy covers approximately 600,000 students from pre-school (2-K and 3-K) through eighth grade, restricting tools for children up to 14 years old in the nation's largest school district. Under the directive, the city will remove or disable artificial intelligence features from about 40 previously authorized educational programs that do not meet updated safety standards. Companion chatbots designed to simulate realistic conversations or offer mental and emotional support are banned across all grade levels. Teachers remain permitted to use artificial intelligence tools for administrative duties, including lesson planning, scheduling, translation, and communication, but cannot use them for grading or student counselling.

Educational philosophy and local reactions

City officials justified the restriction by stating that foundational learning requires direct human engagement rather than automated software. Mamdani noted at a Brooklyn press conference that tech companies have promoted classroom artificial intelligence without independent evidence of academic benefit for younger students.

Children need teachers and human connection in order to learn and grow. They need to develop skills alongside their peers, build relationships with educators and wrestle with tough problems on their own.

NYC Schools Chancellor Kamar Samuels stated that early education must prioritize human interaction, discussion, and creativity over outsourcing difficult tasks to machines. The parent and educator coalition AIM criticized the one-year ban as too short, having advocated for a two-year moratorium, and argued that experimental school pilots treat students as test subjects. Randi Weingarten, president of the American Federation of Teachers, endorsed the direction, noting that New York's policy aligns with proposals from the union.

High school pilots and device restrictions

High school students in grades 9 through 12 are exempt from the blanket ban but face specific limits. The district is introducing twice-yearly, 45-minute critical thinking modules for all high schoolers, covering topics such as algorithmic bias, ethical risks, and career impacts. Up to 50,000 high school pupils, representing roughly 5% of the student body, will participate in five supervised pilot programs capped at five classes per school. These pilots include Quill for reading comprehension, Edia for mathematics, Brisk Teaching for teacher-directed assignments (limited to 10 to 20 minutes weekly), Playlab for evaluating output bias, and Intel AI-Ready Schools for semester projects. Alongside the software restrictions, the district recommended daily digital device limits for students.

National and international regulatory context

New York City's policy represents the strictest local measure in the United States, where federal legislation on classroom artificial intelligence has stalled. States including Virginia, Idaho, and Oklahoma have instead directed agencies to issue privacy and security guidance, while Los Angeles, the second-largest US school district, maintains temporary tool blocks and screen limits established in June 2026. Internationally, Norway barred generative artificial intelligence in primary grades one through seven in June 2026, a policy that took effect at the end of August 2026. In Europe, Italy enacted Law 132/2025 in October 2025 requiring parental consent for users under 14, while the European Union AI Act has set a compliance deadline of 2 December 2027 for Annex III high-risk educational regulations.