New York City Schools Bar Student AI Use Through 8th Grade

TL;DR

Starting this fall, New York City’s public schools will stop letting students in prekindergarten through 8th grade use generative AI — software that produces text or conversation on its own, including chatbot “companions” — for one full school year. It covers roughly two-thirds of the city’s public school students. High schoolers are not banned outright: they get limited, closely watched pilot programs and lessons on how to use AI critically. Teachers can still use AI tools for planning and admin work. A new oversight group will study the pause during the year and recommend whether to keep it.

What happened

New York City Public Schools will impose a one-year moratorium — a temporary halt — on student-facing generative AI, including companion chatbots, for students from pre-K through 8th grade, starting in the 2026-27 school year. The restriction will affect nearly 600,000 students, roughly two-thirds of the city’s public school enrollment.

The rule targets student use, not staff use: teachers can keep using AI for lesson planning and other administrative tasks, as long as the tools meet the district’s safety standards. According to Chalkbeat New York, approved AI is limited to lesson planning, translation and drafting communications — not grading, behavior monitoring, counseling or special-education planning.

The policy also sets recommended daily screen-time caps for younger grades: restricted one-to-one device use for the youngest students, about 30 minutes for grades 3-5, and about 45 minutes for grades 6-8. It carves out exceptions for assistive technology for students with disabilities, tools for multilingual and English-language learners, and career-readiness programs such as computer science courses, and, per Chalkbeat, for standardized reading and math tests taken on computers.

High school is treated differently: rather than a ban, Chalkbeat reports students there will get a limited slate of AI tools through closely monitored pilot programs, plus lessons on AI literacy and critical thinking.

Chancellor Kamar Samuels framed the policy this way: “Innovation does not mean more technology, and technology serves learning—not the other way around.” Mayor Zohran Mamdani added, “Children need teachers and human connection in order to learn and grow.” The Education Department told Chalkbeat that “Younger students need hands-on learning, social interaction, play, and teacher-guided instruction. Routine individual screen use can crowd out those experiences.”

A new Technology in Schools Coalition of educators, parents, students and other stakeholders will convene during the 2026-27 school year to evaluate the moratorium and recommend whether it continues.

The move follows months of pressure. In March 2026, the department had already issued preliminary guidance using a three-tier, stoplight-style system: green for approved uses like brainstorming lesson plans, yellow for uses needing professional review, and red for uses like grading or discipline decisions that stayed off-limits. At the time, Samuels said, “AI is here, and our responsibility is to put strong systemwide safeguards in place.” Then, in a June 9, 2026 letter, 29 City Council members — more than half the council — urged Mamdani and Samuels to go further, calling for a two-year moratorium and arguing the existing guidance had no proposals to strengthen student data-privacy protections around AI companies. K-12 Dive reports the letter cited risks to students’ cognitive development, creativity, mental health and the environment, as well as an April 2026 state comptroller audit that found the district’s policy did not fully align with cybersecurity standards.

A national poll adds context, though it is not NYC-specific. Ipsos surveyed 545 K-12 teachers nationwide between April 27 and May 5, 2026, using a household panel recruited by random sampling rather than sign-up, with a margin of error of plus or minus 5.0 percentage points. It found 54% of teachers said AI is making it harder for students to learn critical thinking, and 55% said it’s mostly a shortcut for students to avoid work. The same poll found 74% of teachers believe AI has bigger implications for K-12 education than past shifts like the internet or computers.

What this means (and what it does not)

New York is not moving alone. Los Angeles Unified, the nation’s second-largest school district, separately bars generative AI on district-issued devices and is phasing in its own screen-time limits starting the same school year, though its rule is described as applying to district devices specifically, not the broader restriction New York has announced. More broadly, at least 37 states already have official AI guidance schools can draw on, even though, per CNN, there is no single agreed definition of AI literacy or how to teach it.

The parties around this announcement have different stakes in how it reads. The mayor and chancellor who announced it get to present themselves as putting child development ahead of technology adoption. The council members who spent months pushing for a two-year freeze got a shorter, one-year version of what they asked for. Ipsos, whose poll is the empirical backing most directly cited in the reporting reviewed here, sees its findings amplified. None of that makes the underlying figures wrong — it is simply useful to know who is invested in which reading of them.

What the policy does not do is settle the underlying debate with New York-specific evidence. The official city announcement does not cite specific data, research studies, or documented incidents of harm in NYC schools; it is framed as precautionary. The teacher survey behind much of the public debate is national, not local, and the comptroller audit council members cited concerned cybersecurity and data-privacy gaps, not classroom harm. A reader should not conclude the city has documented AI hurting its own students — that evidence, if it exists, is not in the materials reviewed for this article.

What we still do not know

It is not established what documented incidents of harm, if any, occurred in NYC schools specifically — the poll and the audit cited as backing are national and state-level, not local evidence. Nor is it clear what metrics the new Technology in Schools Coalition will use to judge success, or what would justify extending the moratorium past one year versus letting it lapse. How the rule will be enforced on students’ personal devices or home internet access, outside school-issued equipment and networks, is not addressed in the sources reviewed, and it is not clear whether the ban covers AI features already built into approved software — such as spell-check or adaptive learning platforms — or only standalone chatbots and generative tools. The consequences for schools, teachers or vendors who violate the moratorium are not specified. How New York’s rule compares in scope to Los Angeles’s — whether it is limited to school-issued devices and networks, as LAUSD’s is, or extends further — is unclear from the reporting reviewed. And the City Council’s June letter itself could not be independently read for this article; its contents here rely on K-12 Dive’s reporting of it rather than the original document.