Six hundred thousand students. Forty educational tools. One year.
Those are the numbers behind the decision New York City announced in early September 2026, and they are why experts say it could set the stage for how other districts approach AI. The city, whose public school system is the largest in the United States, is ending the classroom use of AI tools for elementary and middle school students for one year. What it does and does not say is easy to get wrong in a headline, so here is a plain account, based on reporting by Al Jazeera (Al Jazeera).
What exactly is being paused?
The moratorium ends the use of 40 different educational tools for students through eighth grade, affecting about 600,000 students. It runs for one year, during which the city says it will study the effect of the policy on students. Mayor Zohran Mamdani framed it in human terms: "Children need teachers and human connection in order to learn and grow," he said, adding that they need to build relationships with educators and "wrestle with tough problems on their own."
The policy also sets screen-time limits, with staggered restrictions that limit use to less than 45 minutes each day through eighth grade, and bans "companion chatbots" through high school. City Hall did not elaborate, when asked, on what that ban covers.
What is not covered?
Several things are deliberately left out, and they matter as much as the ban itself. Teachers are not barred from using AI to create lesson plans, so the pause is aimed at what students use in class, not what teachers use behind the scenes. There are exceptions for students with disabilities and for "multilingual learners," and for environments such as "career readiness programs, like computer science," where students are likely to use generative AI in ways that fit the course. Students will also be given access to two yearly AI literacy courses, designed to help them understand the technology as well as bias and ethical issues.
The city is also running a pilot in high schools for a limited number of students. Among the tools being tested is Edia, a maths platform that provides "step-by-step coaching" and prompts students to explain their reasoning, limited to 20 minutes a week. Another, Quill, will be used in English language arts and limited to 15 minutes a week. In other words, the policy is not a rejection of every AI tool. The pilot points to a model of a short list of named tools with strict weekly time limits.
What is the context?
Al Jazeera places the decision against a difficult backdrop. Reading proficiency scores among students in grades 3 through 8 fell by 6 percent in 2025, and the article says the moratorium "comes at a consequential time" for the city's schools. It also points to a review by the Brookings Institution's Center for Universal Education of 400 studies, along with interviews with teachers, parents and other educators, which concluded that the negatives of generative AI in the classroom outweigh the positives. The same report stressed that careful and limited implementation in specific use cases can help enrich learning outcomes.
One theme in that review was what its authors called a "digitally induced 'amnesia,'" in which students were unable to recall basic information about topics they had studied or submitted work on. One legal expert quoted in the article said the important change is in the default: "the products have to be proven safe before they can be used by students."
What are the doubts?
Not everyone is convinced it will work. Murray, an associate professor of communication at the University of Michigan-Dearborn, told Al Jazeera that "it is hard to know the effectiveness of the ban because they haven't shared the enforcement mechanism." And a school can regulate what happens in class far more easily than what happens at home. Asked about homework, the city's schools chancellor, Kamar Samuels, said the policy seeks to limit AI use and to help teachers understand how to combat its use outside school. The article does not report a specific way of enforcing that at home.
So the policy is best read as a serious experiment, with a built-in year of study, and not as a settled verdict on AI in schools.
What can other districts take from it?
You do not need to ban anything to learn from New York's approach. Three ideas travel well. The first is to flip the default so that a tool has to be shown to be safe and useful before students use it, instead of being allowed until a problem appears. Our procurement scorecard and privacy checklist give you a way to do that. The second is to set time limits and purposes for any tool, as the pilot does with its 15 and 20 minutes a week, so that use is deliberate. The third is to pair any restriction with teaching. New York is adding AI literacy courses alongside the pause, which matches our own advice on teaching students to use AI responsibly.
For a school without a district-wide policy, the takeaway is more modest but just as real. Students meet AI in every classroom, and when teachers get no guidance they make different decisions. Our guide to writing a school AI policy students can follow includes a template you can adapt.
Frequently asked questions
Is AI banned in all New York City schools? No. The moratorium covers educational AI tools for students through eighth grade for one year, with exceptions and a limited high-school pilot. Teachers can still use AI to prepare lessons.
How long does it last? One year, during which the city says it will study the effects.
Does it affect homework? A school can regulate class time far more easily than home use. The chancellor said the policy aims to limit AI use and to help teachers respond to outside use, and the article reports no specific enforcement method.
What happens next? The city will study the policy over the year and run its high-school pilot. Other districts will be watching the results.
Related guides
- How to write a school AI policy
- Teaching students AI literacy
- The procurement scorecard
- Education News: latest research, a live feed of new education studies