Abstract:
The New York City Department of Education announced that it will implement a one-year restriction on the use of generative AI in public schools during the 2026-2027 school year, prohibiting students from preschool to eighth grade students from using generative AI tools for students. As one of the largest public school systems in the United States, New York City hopes to reduce the reliance on generative AI in classrooms and encourage students to develop the ability to think independently, solve problems and question information. The measure is not a permanent ban or a statewide policy.
Under the new rules, high school students can only use generative AI in approved pilot projects or career preparation programs overseen by teachers. At the same time, all high school students will receive basic AI literacy courses twice a year. Regardless of the grade a student is in, AI chatbots designed for companionship and emotional interaction are prohibited.
The new rules also limit screen time in class for younger students: a maximum of 45 minutes per class for middle school students and a maximum of 30 minutes for elementary school students. However, accessibility tools, exam assessments, remote learning, and approved programming and robotics activities are exempt from these restrictions.
Teachers are also subject to the new regulations. They can use approved AI tools to assist with lesson preparation, but they may not use them for grading, behavior monitoring or making important decisions about students. The New York City Department of Education said the restrictions are intended to protect human connection, curiosity and creativity in schooling.
New York City Schools Superintendent Kamal Samuels said schools need to set necessary boundaries for the use of AI to protect the critical human connections, curiosity and creativity that help children grow. New York City Mayor Zoran Mamdani said that schools teach children every day to ask questions, question assumptions and ask "why". When AI enters schools, the education department also has the responsibility to adopt this attitude.
This is not the first time New York City has tried to restrict students’ use of AI. In early 2023, New York City public schools banned students from using ChatGPT, but later revoked this restriction. At the time, education officials argued that the initial response placed too much emphasis on risks such as cheating, misinformation and weakened critical thinking.
As generative AI becomes more prevalent in education, concerns are growing about students’ overreliance on such tools. Some studies have pointed out that in the absence of supervision, AI that can directly provide answers may weaken students' ability to learn independently and pose emotional and trust risks. Student cheating has been a particular concern, and schools and teachers have tried to deal with it through handwritten assignments, oral exams, and designing questions that are difficult to directly hand over to large language models. Some professors even use hidden prompt words to set up "traps" to test whether students directly rely on AI to complete assignments.
New York City’s policy is not alone. The Los Angeles Unified School District is also currently temporarily restricting access to generative AI tools on school devices for all grades.
The New York City Department of Education will form a coalition to evaluate this policy and related pilot projects and make recommendations in April 2027. The recommendations will help decide whether to continue existing policies next school year or adjust how AI is used in schools.
Comments