The nation's broadest school AI moratorium
New York City public school students returned to classrooms this September under one of the most dramatic education technology policies ever attempted in the United States: a one-year moratorium on student-facing generative AI, stretching from 2-K — the city's free early care and education program for two-year-olds — all the way through eighth grade. Mayor Zohran Mamdani, who unveiled the policy on September 2, less than a week and a half before the September 10 start of the school year, described it as the broadest generative AI moratorium in the nation.
The restriction applies to student-facing generative AI — software that creates text, images, audio, or other content in response to a student's instructions. It is not a ban on every technology that uses AI. Students can continue using assistive technology required under an individualized education program or Section 504 plan, and the city's guidance preserves necessary tools for students with disabilities and English-language learners. Career-readiness programs also have a pathway to use certain AI tools where they are needed for relevant coursework. But the general-purpose chatbots that have become a fixture of homework and classroom life are out.
"Companion chatbots will be prohibited across all grades," according to NYC Public Schools' policy for the 2026–27 school year. The city is also reviewing technology already in use: officials said more than 38 previously allowed programs would have their AI components disabled or be discontinued where those components cannot be switched off. Products from Amira and HMH were specifically discussed during the September 2 announcement.
What each grade tier gets
The policy is tiered by age, reflecting the city's judgment that younger students need the most protection:
- Kindergarten through grade 2: generative AI banned outright, with zero minutes of one-to-one device use recommended. The only exception is assistive technology required under an IEP or 504 plan.
- Grades 3–5: generative AI banned, with a recommended maximum of 30 minutes of daily one-to-one screen time. Exceptions cover IEP and 504 plans plus multilingual learner tools.
- Grades 6–8: generative AI banned, with a recommended maximum of 45 minutes of daily screen time. Exceptions add specialized career preparation.
- Grades 9–12: controlled access through pilot programs and career tracks. Every high school student must complete two 45-minute AI literacy modules covering how AI works, data privacy and bias, and appropriate uses of AI in school.
For high schools, the city has centrally approved five pilot programs — Quill, Edia, Brisk Teaching, Playlab, and Intel AI-Ready Schools — selected, the city says, using safety and data-privacy standards. The pilots do not introduce general-purpose chatbots such as ChatGPT or Claude into the curriculum; they are bounded tools aimed at specific subjects and teaching tasks. Career and technical education programs have a separate pathway for students who need AI tools to develop job-related skills.
Teachers, meanwhile, can still use AI for tasks like lesson planning, but they are not allowed to use it for grading, placement, or other decisions about students — a line drawn to protect students from automated judgments about their abilities and futures.
The argument for the ban
The philosophy behind the moratorium is that students need to learn to think for themselves before handing intellectual work over to machines. Education experts who support the policy's central premise argue that foundational skills — reading, writing, reasoning, arithmetic — must be built through genuine cognitive effort before students gain access to tools that can do that work for them.
CUNY education experts Professor Stephen Farenga and Queens College Professor Salvatore Garofalo, who study AI and cognition and have a forthcoming book on human cognition in the age of AI, draw a distinction between "cognitive offloading" and "cognitive surrender." Offloading is ancient and benign: a notebook saves you from having to remember everything, a calculator from having to do every sum by hand. Surrender is what happens when students hand over critical thinking entirely to the AI. The city's bet is that the moratorium protects younger students from surrender while their thinking skills are still forming.
The policy also follows last year's statewide cellphone ban, which prohibits personal internet-enabled devices — cellphones, laptops, tablets, portable music devices — throughout the school day. Together, the two policies represent a striking reversal of the decade-long push to put a screen in every student's hands.
The case against: bans, the digital divide, and forbidden fruit
Not everyone is convinced a moratorium is the right tool. Farenga himself questions whether an outright ban accomplishes the goal: "When you go to ban something outright, it almost makes it more desirable for people to want to look at and to want to get into," he said, warning that failing to teach students how to use AI could itself be dangerous given how deeply the technology is already embedded in society.
Equity advocates raise a sharper concern: will the ban exacerbate the city's digital divide? Students in well-resourced households will have AI at home regardless of what happens at school. Students who rely on school for their technology exposure get a year of AI literacy modules in high school and nothing before that. "The idea that you can learn AI fluency without using AI, I analogize this to trying to learn the piano by reading music," said David Adams, CEO of the Urban Assembly, a nonprofit that supports more than 20 public schools in New York City. "This kind of abstract reasoning, knowledge-based understanding of AI will never be as good as playing Für Elise and having to practice that and get that under your fingers."
There is also the enforcement question. A ban on school devices and school networks does not reach personal phones — although the statewide cellphone ban covers the school day — and students will encounter AI on home devices, at libraries, and in the apps they already use. The city can control its classrooms; it cannot control the technology's presence in children's lives.
A national test case
What makes the New York policy significant is its scale and its clarity. Other districts have written rules about AI — districts in Florida and elsewhere have published guidance on disclosure and assessment integrity this September — but the trend has been toward managed use, not prohibition. New York City, operating the largest school system in the country, has gone the other way, and every other district in America will be watching what happens.
The experiment is time-bound: a one-year moratorium. That gives the city a natural evaluation point. If reading and math scores improve, if teachers report deeper engagement, if parents are satisfied, the moratorium could become a model — or be extended. If students fall behind in AI fluency while their peers elsewhere learn to work with the technology, the backlash will be swift.
For now, the city's youngest students are doing something their older siblings never did: learning to read, write, and reason with no generative AI in the room at all. In a year, the data will start to tell us whether that was protection or deprivation.
Sources
