New York City AI School Ban Explained

New York City is imposing some of the nation’s strictest restrictions on artificial intelligence in public schools for the 2026-27 school year. Students from 2-K through eighth grade will generally be prohibited from using generative AI tools including chatbots and AI tutors, affecting nearly 600,000 children. The city is also restricting individual screen time, banning AI companion chatbots across all grades and prohibiting teachers from using artificial intelligence to grade student work. High school students will be allowed limited AI use, including tightly controlled pilot programs and AI literacy instruction. The policy represents a major reversal from preliminary guidelines issued in March 2026 that generated intense backlash from parents, teachers and elected officials who argued that New York City was embracing AI too quickly.

AI Banned Through Eighth Grade

New York City will prohibit student-facing generative AI from 2-K through eighth grade. The restriction covers children in early-childhood programs, elementary schools and middle schools and includes tools such as generative AI chatbots and AI tutors. [Chalkbeat] [News 12]

Nearly 600,000 students are affected by the restriction. That represents roughly two-thirds of students in New York City’s public-school system. [NYT] [News 12]

The policy applies during the 2026-27 school year. City officials characterized the restrictions as a one-year moratorium while New York continues studying the effects of generative AI on children and develops longer-term rules. [News 12]

Students will not simply be allowed unrestricted AI access once they reach high school. High school use will remain limited to approved educational purposes, particular programs and controlled experiments rather than allowing students to freely use consumer AI tools throughout the school day. [Chalkbeat] [NYT]

ChatGPT And Other AI Tools

The restrictions broadly target generative AI rather than only ChatGPT. The policy covers artificial-intelligence systems capable of generating text, images or other material in response to student prompts, including chatbots and AI tutoring systems. [Chalkbeat]

The rules therefore should not be understood simply as a ChatGPT ban. ChatGPT is one prominent example of the technology affected, but the policy is designed to govern the broader category of student-facing generative AI regardless of the company that develops the tool. [Chalkbeat]

AI embedded inside ordinary educational software creates a complication for enforcement. Artificial intelligence is increasingly built into software schools have already been using for years, making it difficult to separate a dedicated chatbot from AI functions added to conventional educational products. [Chalkbeat]

AI Companion Chatbots Banned In All Grades

AI companion chatbots will be prohibited for students regardless of age. Unlike the broader generative-AI moratorium, which primarily covers students through eighth grade, the prohibition on chatbots designed to provide companionship or emotional support applies across all grade levels. [NYT] [News 12]

The restriction reflects concerns about students treating artificial intelligence as a substitute for human emotional support. New York City officials have emphasized that technology should support rather than replace relationships between students, teachers, counselors and other adults. [NYC Public Schools]

Screen Time Limits

The AI policy is accompanied by unusually strict limits on individual screen use. New York City’s new technology framework goes beyond regulating artificial intelligence and addresses how frequently students should be using laptops, tablets and other personal screens during the school day. [NYT] [Chalkbeat]

Students from 2-K through second grade generally will not receive individual screens for classroom instruction. Teachers can still use shared classroom technologies such as interactive smart boards, and exceptions are available for certain educational needs. [Chalkbeat]

Students in grades three through five are recommended to spend no more than 30 minutes per day on individual devices. The limit is intended to preserve more classroom time for teacher-led instruction, reading, writing, conversation and hands-on activities. [Chalkbeat] [News 12]

Middle school students are recommended to receive no more than 45 minutes of individual screen time per day. The recommendation applies to students in grades six through eight, although exceptions can apply to certain programs and educational needs. [Chalkbeat] [News 12]

High schools will have greater flexibility over screen use. Recommendations for older students can vary depending on the subject, instructional approach and purpose of the technology. [Chalkbeat]

Why New York City Is Limiting Screens

Education officials say younger children particularly need learning experiences that do not take place through screens. A city presentation obtained by Chalkbeat said younger students need hands-on learning, social interaction, play and teacher-guided instruction and warned that routine individual screen use can displace those experiences. [Chalkbeat]

The restrictions reflect a broader debate about “cognitive offloading.” Education officials and critics of classroom AI have expressed concern that students may begin delegating writing, reasoning, research and problem-solving to artificial intelligence rather than developing those abilities themselves. [NYC Public Schools] [Guardian]

Teachers Can Still Use AI

The policy does not ban teachers from using artificial intelligence. Educators will continue to be permitted to use approved AI systems for certain professional and administrative tasks. [Chalkbeat]

Teachers can use approved AI tools to help prepare lesson plans. New York City says technology can be used to improve efficiency as long as educators remain responsible for the substance and quality of instruction. [NYC Public Schools] [Chalkbeat]

Teachers may also use AI for translation and drafting communications. These uses are considered lower-risk because the technology assists educators rather than making consequential decisions about students. [Chalkbeat]

Teachers Cannot Use AI To Grade Students

Teachers are prohibited from handing grading decisions over to AI. The educator responsible for the class must determine what a student knows rather than relying on an artificial-intelligence system to assign a grade. [NYT] [NYC Public Schools]

The restriction predates the September student ban. New York City’s March AI framework had already placed assessments and grading in its highest-risk “red” category and said AI-generated information could only be advisory. [NYC Public Schools]

Concerns about AI grading had grown nationally before the policy was finalized. An August 2026 survey cited by NBC New York found that roughly one-quarter of teachers surveyed said they had used AI to grade or provide feedback on student assignments or assessments. [NBC New York]

Other AI Uses Teachers Cannot Delegate

AI cannot make important decisions about individual students. New York City’s policy prohibits artificial intelligence from determining placement, discipline, eligibility, promotion, graduation or access to programs. [NYC Public Schools]

AI cannot write IEP or Section 504 plans for students with disabilities. Those plans must be developed by qualified professionals rather than generated by an automated system. [NYC Public Schools]

AI cannot be used for behavioral surveillance of students. The city’s framework classifies behavioral monitoring and student surveillance among prohibited uses. [NYC Public Schools]

Teachers also cannot use AI as a substitute for counseling students. The September policy identifies counseling and emotional-support functions as activities that must remain centered on human professionals. [Chalkbeat]

High School AI Pilot Programs

A limited number of high school students will still be able to use generative AI. Rather than imposing a complete systemwide prohibition, the city plans controlled pilot programs designed to determine whether AI can produce educational benefits for older students. [News 12]

No more than approximately 50,000 high school students are expected to participate in the pilot programs. That represents about 5 percent of the overall New York City school system. [News 12]

Potential high school uses include literacy and career preparation. Education officials argue that older students need some guided experience with artificial intelligence because they are entering colleges and workplaces where the technology is increasingly common. [Chalkbeat]

High school students will also receive AI literacy instruction. The city plans twice-yearly lessons focused on artificial intelligence, critical thinking and understanding the risks and limitations of AI systems. [News 12]

Exceptions To The AI Ban

The prohibition contains several exceptions. Approved instructional tools can remain available even when they incorporate some artificial-intelligence functions. [NYT] [Chalkbeat]

E-books, coding tools, robotics and educational simulations may be exempt. The city can centrally approve programs that serve an instructional purpose even if some underlying features incorporate AI. [Chalkbeat]

Computer-based tests and diagnostic assessments can also be exempted. Schools already rely on digital testing platforms for required assessments in subjects including reading and mathematics. [Chalkbeat]

Students with disabilities can receive technology exceptions. Assistive technology necessary to provide appropriate educational access will not be eliminated merely because it incorporates AI or requires additional screen time. [Chalkbeat]

Multilingual students may also receive exemptions. Tools intended to support students learning English can be permitted when education officials determine that they are appropriate. [News 12]

New York City Originally Banned ChatGPT In 2023

New York City previously blocked ChatGPT on school networks shortly after the chatbot became popular. In January 2023, the Education Department restricted access because of concerns about cheating and the potential negative effect on student learning. [Chalkbeat]

The city reversed that approach only months later. Then-Schools Chancellor David Banks argued that simply blocking the technology would not prepare students for a world in which generative AI was becoming widespread and said schools should instead teach students how to use it responsibly. [Chalkbeat]

The 2026 policy represents another major shift in New York City’s approach. Rather than attempting to either completely block AI or broadly integrate it, the new rules establish different levels of access depending on student age and the educational purpose involved. [Chalkbeat]

March 2026 AI Guidelines

New York City released preliminary AI guidance in March 2026. The original framework was designed around a traffic-light system separating prohibited, cautionary and lower-risk uses of artificial intelligence. [NYC Public Schools]

The guidance emphasized responsible AI adoption rather than a student ban. Officials said the goal was to help educators use artificial intelligence safely and ethically while preserving professional judgment and student privacy. [NYC Public Schools]

Student AI use received comparatively little attention in the original version. That omission became one of the central complaints from parents, educators and elected officials who believed the framework addressed how teachers could use AI without adequately considering whether young children should use it at all. [Chalkbeat]

Parent Backlash

The March guidelines prompted a major parent backlash. Parents packed education meetings and argued that the city was moving forward with classroom AI without enough evidence that generative technology benefited children. [Chalkbeat] [Guardian]

Parents for AI Caution in Educational Spaces became one of the leading groups pushing for restrictions. Brooklyn parent Kelly Clancy helped organize families calling for a temporary halt to classroom generative AI while researchers and educators studied the consequences. [NBC New York]

Thousands of people signed a petition calling for a two-year AI moratorium. The movement argued that schools should establish evidence of educational value and effective safety rules before introducing generative AI to children. [Chalkbeat]

City Council Pushes For A Moratorium

Twenty-nine New York City Council members called for an immediate pause on generative AI in schools. The lawmakers signed a June letter to Mayor Zohran Mamdani urging the administration to impose a two-year moratorium. [NY1]

Council members cited concerns involving learning, privacy and equity. Supporters argued that schools should not experiment on children with rapidly evolving technology whose long-term developmental effects are still uncertain. [NY1]

Chancellor Kamar Samuels Says City “Missed The Mark”

Schools Chancellor Kamar Samuels acknowledged that the original AI rollout had failed to adequately address community concerns. During the backlash, Samuels said the Education Department had “missed the mark” and promised stronger guardrails, particularly for younger students. [Chalkbeat]

The city delayed its more comprehensive AI policy. Officials originally expected to release a fuller AI Playbook in June but postponed the rollout while reconsidering the policy following parent and educator opposition. [Chalkbeat]

Software Purchasing Freeze

New York City temporarily froze new school software purchases while rewriting its AI rules. Samuels instructed principals in July to pause purchases so officials could determine which programs complied with the new technology framework. [Chalkbeat]

The freeze caused operational problems for some principals. School leaders told Chalkbeat they were unable to renew ordinary programs including electronic gradebooks, digital hall passes and family-communication systems even when those products were not primarily AI tools. [Chalkbeat]

The confusion demonstrated how difficult separating AI from ordinary school technology has become. Artificial-intelligence functions are increasingly being incorporated into software that schools purchased long before generative AI became a major education controversy. [Chalkbeat]

Some Parents Say The New Ban Does Not Go Far Enough

Parents who demanded a full two-year moratorium have said the new rules remain too permissive. Clancy praised the pressure parents placed on the city but expressed concern that exemptions for centrally approved software could allow AI to remain much more common than the policy’s headline ban suggests. [NYT]

Critics are particularly concerned about AI hidden inside approved instructional programs. A program may contain generative or adaptive AI even when students do not experience it as a conventional chatbot, making enforcement and transparency more complicated. [Chalkbeat]

Teachers Question Whether An AI Ban Can Be Enforced

United Federation of Teachers President Michael Mulgrew questioned how schools could completely control AI use. He noted that students already use artificial intelligence outside school and that AI is increasingly embedded inside commonly used software. [Chalkbeat]

Enforcement remains one of the policy’s unresolved challenges. Schools can restrict approved software and devices on their networks, but they have far less control over AI tools students access on personal devices or outside school hours. [Chalkbeat]

AI Task Force And Future Policy

New York City plans to continue reviewing the restrictions during the school year. A coalition involving students, educators, parents, elected officials, unions and technology experts will examine how the policy works in practice. [News 12]

The 2026-27 restrictions are not necessarily permanent. Officials intend to use the year’s experience, research and feedback to develop recommendations governing artificial intelligence in future school years. [News 12]

New York City’s policy could influence school districts around the country. As the largest public-school system in the United States, the city’s move from early ChatGPT prohibition to AI experimentation and then to age-based restrictions provides a prominent test of how schools can respond to rapidly changing generative technology. [NYT] [Chalkbeat]