NYC Schools Ban AI Chatbots Through 8th Grade, Ease In High School Use
America's largest school district is pausing generative AI for younger kids for a year while letting some teens try supervised pilot programs — pleasing almost no one completely.
What happened: New York City, the country's largest school district, has announced a one-year moratorium on generative AI for students from pre-K through eighth grade, starting with the school year beginning September 10. Mayor Zohran Mamdani said the city will disable AI features in 38 previously approved apps, ban AI "companion" chatbots at every grade level, and limit young children's screen time and device access. High schoolers avoid a full ban: up to 50,000 can join five supervised pilot programs using pre-vetted tools, alongside new mandatory AI literacy classes. The move landed the same week Los Angeles confirmed its own AI and screen-time restrictions, while San Francisco is still writing its policy.
Why it matters: This is the broadest AI restriction any major U.S. school system has adopted, and other districts are watching closely. It reflects growing unease among parents, teachers, and researchers that AI chatbots and tutoring tools might undercut attention spans and basic skill-building before kids are ready to use the technology critically. But by drawing a hard line at eighth grade rather than banning AI outright, the city is also making a bet: that teenagers need supervised exposure to prepare for a job market that increasingly expects AI literacy, even as younger children are shielded entirely.
How it works, plainly: Pre-K through second graders lose access to individual devices in class; grades three through five get a 30-minute daily screen cap. Generative AI and companion chatbots are switched off system-wide for pre-K through eighth grade. High school pilots use specific vetted tools — Quill for reading, Edia for math, Brisk Teaching and Playlab for teacher-built activities, Intel's AI-Ready Schools for projects — capped at 45 minutes a week under direct teacher supervision. Older students also take two 45-minute AI literacy sessions a year; pilot participants get 12 to 15 hours. Teachers may use AI to help plan lessons but not to grade work.
Who's still unhappy: Enforcement is the biggest gap: Mamdani offered no concrete mechanism, saying he trusts teachers, principals, and superintendents to make it work. The teachers union wants the city to vet AI products before purchase rather than leave schools to sort out compliance later. A parent coalition that wanted a stricter two-year, all-grades moratorium calls this a real step but says high school carve-outs leave students exposed. A teen activist argued protection shouldn't stop at middle school, while an anti-poverty nonprofit leader warned the younger-grade ban could leave kids unprepared and more likely to misuse AI unsupervised at home.