New York City is sharply restricting the use of generative artificial intelligence in its public schools, barring nearly 600,000 younger students from using the technology during the 2026-27 academic year while placing significant limits on its use among high school students.
Mayor Zohran Mamdani and Schools Chancellor Kamar Samuels announced the one-year moratorium Wednesday, days before students are scheduled to return to school. The policy covers children from 2-K through eighth grade, representing roughly two-thirds of the student population in the nation’s largest public school district.
“The tech industry wants us to believe that A.I.-powered early education is not only inevitable, but necessary,” Mamdani said in a statement. “We do not see it that way.”
Students in high school will still be allowed to use artificial intelligence, though only under restricted educational circumstances. Each high school student will receive two 45-minute lessons on AI literacy during the school year, addressing subjects including technological bias, ethics, potential effects on careers and the limitations of artificial intelligence.
The city will also launch a limited trial involving five AI programs that have been reviewed and approved for classroom use. The pilot will be conducted in a small group of high schools and will involve no more than 50,000 students, representing approximately 5% of the city school system’s total enrollment.
One category of artificial intelligence will be banned across the board: companion chatbots that simulate friendship or are marketed as offering emotional support or mental health guidance will not be permitted for students in any grade.
Educators, meanwhile, will retain access to certain approved AI applications. Teachers will be permitted to use the technology to prepare lessons, translate educational material and compose communications, but they will be barred from relying on AI to grade students’ work or make sensitive student-related determinations.
The new rules include several exemptions. AI may continue to be used in instructional programs that have received central approval, for testing purposes and as assistive technology for students with disabilities. Exceptions will also apply to multilingual students and to certain career-preparation programs, including computer science courses.
“We’re putting guardrails in place to protect the critical human connection, curiosity and creativity that help children grow,” said Samuels, who added that the school system would not “conflate innovation with more tech.”
The Mamdani administration is also moving beyond AI to reduce students’ overall exposure to screens in school. Individual screen use will be restricted for children in second grade and younger, while the city is recommending a maximum of 30 minutes per day for third- through fifth-graders and 45 minutes daily for students in grades six through eight.
The tougher approach comes after months of debate over the rapidly expanding role of artificial intelligence in education. In March, the city school system issued preliminary guidelines that permitted responsible AI use while stressing academic honesty, student privacy and the continued importance of human judgment. The city had also blocked ChatGPT from school networks and devices in early 2023, only to lift that restriction later in the year.
“This is certainly a step forward,” said City Council Member Eric Dinowitz, chairman of the council’s Education Committee. “I still have a lot of questions.”
Some parent advocates are similarly questioning how comprehensive the restrictions will prove to be, particularly because AI capabilities are increasingly being incorporated directly into educational software that schools already use.
“What does that mean in a world where all software at this point is now using generative A.I. as part of their core features,” said Naveed Hasan, a member of the city’s education oversight panel. “It remains to be seen.”
