9/16/26
New York City has announced a ban on the use of Artificial Intelligence (AI) in the public school system for the 2026-2027 school year for students in grades K-8. The nation’s largest student AI moratorium to date will block roughly 600,000 students from mainstream tools like ChatGPT and Claude, affecting 38 previously approved educational programs. The policy is a small step towards regulating the explosive growth of AI usage in education. Before ChatGPT’s release in 2022, generative AI in education was almost non-existent. Since then, it has grown to a $760 million industry and is projected to grow over 40% annually.
Announcing the moratorium policy in September 2026, NYC Mayor Zohran Mamdani highlights the importance of students developing independent problem-solving skills, arguing that children need to “wrestle with tough problems on their own” rather than offloading their critical thinking onto others. The main concern with the use of AI in education is that they remove the cognitive friction of solving a problem, which is the primary driver of learning. Mayor Mamdani also reiterates the importance of student-educator relationships and the human aspect of the school system, which AI cannot provide, such as the opportunity to grow and learn alongside peers.
The newly implemented moratorium follows previously unsuccessful attempts by the NYC Department of Education to address major concerns about the rising use of AI in education. In March of 2026, the NYC public schools released an AI usage guidance that quickly faced backlash for schools’ lack of substantive action. The guidance also acknowledges only educators’ use of AI while glossing over student use. In response to this, disappointed parent groups organized and pushed for a two-year moratorium on AI use across all grades. The New York Department of Education eventually acknowledged this moratorium after parents’ continued push through the spring 2026.
Proponents of AI in education believe that banning AI tools in schools is not the solution. Jean-Claude Brizard, CEO of Digital Promise, an education innovation and technology non-profit organization, argues that the NYC schools’ efforts are misguided because “it’s like asking someone to remove added sugar in your food – it is everywhere; it is pervasive.” Instead of resisting, proponents of AI in education argue that it is better to adapt and harness AI’s “immense potential to advance educational equity” when used responsibly. Massive AI companies such as Microsoft and OpenAI have also pledged millions of dollars to support the American Federation of Teachers in providing AI training for 400,000 teachers in the union. Regarding such expansion of big tech into education, Mayor Zohran Mamdani criticized the industry, saying: “The tech industry wants us to believe that AI in early education is not only inevitable, but that it is necessary. We do not see it that way.”
Studies have revealed the effects of using AI as an educational tool. Most notably, in a Harvard study conducted in 2025, students enrolled in a Harvard physics class using a specialized AI tutor learned roughly twice as much in less time as the gold-standard in-class learning, while reporting higher engagement and motivation. However, the specific study is not generalizable to AI tools in general. The researchers explicitly used an AI tutor custom-engineered with best educational practices in mind, not a base commercial chatbot, in a controlled educational setting. In contrast, a 2025 PNAS study found that students practicing with an unrestricted commercial chatbot scored 17% worse on independent exams than their unassisted peers, as bypassing the “productive struggle” of problem-solving undermined their foundational learning and created a false sense of mastery. Yet this has not stopped profit-oriented AI startups and marketers from citing the Harvard study to promote their own products.
There remains concern about the developmental harm of unregulated AI use in an uncontrolled learning environment, such as public schools. As the researchers around the world continue to study the effects of AI use in the classroom, perhaps the only way to ensure healthy educational outcomes is to heavily regulate generative AI and its integration into early education. Undoubtedly, AI tutors and technologies in general have immense potential in education when used in specialized, controlled settings. However, educational boards across the country must not prioritize technological advancement at the expense of students’ cognitive and emotional development in public school systems.