The nation’s largest school district now has the strictest guidelines for allowing younger students to use artificial intelligence tools in school (at least temporarily). Some education experts say this is a step in the right direction for schools to effectively utilize AI to benefit children.
New York City schools will ban the use of generative AI tools by elementary and middle school students during the next school year, Mayor Zoran Mamdani announced Wednesday. Mamdani said in a statement that the one-year suspension, which affects more than 600,000 students, will give the city’s education authorities time to “study the impact of this technology.”
The announcement comes amid a swirling debate about the role AI should play in how students across the country learn, and how parents and educators can best position children for future success. Some parent-led groups have raised concerns that AI tools could make it easier for students to cheat or make them over-reliant on technology at the expense of creative thinking skills. Advocates of AI in the classroom say the technology is becoming pervasive in modern society, and students will benefit from learning how to use it effectively and responsibly sooner rather than later.
New York state’s new guidelines already have critics on both sides of the debate. Some critics argue that the city is denying students valuable learning tools and the ability to develop AI literacy. Some say the city’s new policy doesn’t go far enough to ban AI tools. According to the New York Times, students will reportedly still be allowed access to “centrally approved educational programs,” with high schools largely excluded.
New York’s new policy initially makes some sense for a middle class of parents who want their children to learn how to use AI responsibly without relying too heavily on it, said Bree Dussault, president and managing director of the Center for Reinventing Public Education, an Arizona State University-affiliated research organization that studies AI implementation in school districts across the country.
The temporary and limited moratorium “represents a clear but flexible policy approach that is necessary in this kind of rapidly evolving moment,” Dussault said. She points out that ultimate success will depend on what city officials discover as they study the impact of educational AI tools over the next year. The agency’s findings, at least in theory, could also help parents across the country who want to strike the right balance between their children and AI education at home, she added.
Other school districts around the country are probably taking notice, too. After New York City and Los Angeles blocked access to ChatGPT in schools in 2023, a series of other cities followed suit. Notably, New York State ultimately reversed its decision. According to an October 2025 study by the College Board, nearly 40% of U.S. schools prohibit the use of AI tools in the classroom, with many schools ultimately leaving the decision up to individual teachers.
Potential “home run”
New York City’s new policy applies to “all software that uses generative AI for students,” according to a city press release. Students are also prohibited from using individually vetted devices until third grade, and students from age 2 to high school are also prohibited from using companion chatbots that simulate emotional connections and friendships. Teachers will be able to use AI tools to create lesson plans and complete administrative tasks, but they will not be able to use them to grade assignments.
The moratorium balances the city’s strategic investments in AI literacy classes for high school students and multiple pilot programs to test AI educational tools in supervised environments with a limited number of high school students, Dussault said.
Dr. Nicole Burns, chief psychologist for education at the American Psychological Association, said the policy could be effective and even a “home run” if cities spend the next year testing and identifying AI educational tools with high data privacy standards that not only capture students’ attention, but also support student learning and critical thinking.
Education technology products built on generative AI often make unsubstantiated claims that they improve learning and critical thinking skills in young students, according to an APA report co-authored by Burns released Thursday. Features that capture children’s attention and increase engagement are often misunderstood as promoting real learning, the report found.
Effective tools must have a track record of helping students “retain and apply knowledge” and maintain human interaction with teachers and peers, which APA research has found is essential to supporting effective learning, Burns said. She is also an advocate of teaching parents and families about AI literacy, helping children understand the pros and cons of the technology and how to use it safely and effectively. Children don’t need “unlimited access to AI” to learn AI literacy, she says.
Her recommendations for parents and teachers: Start educating children as early as possible, including teaching them that “AI can be wrong” and that critical thinking skills are still best demonstrated by humans. She suggests monitoring the tools they use (if they use them at all) and their interactions with those tools to make sure they’re retaining the information and not just being entertained by the technology.
During that time, Burns praises New York for “having the courage to slow down, ask the hard questions, and take on the difficult work.” “Part of that is asking for evidence that people are actually benefiting from the AI that’s being touted.”
