New York City’s AI Ban Is the Wrong Direction for Its Students

New York City has announced the broadest student-facing AI moratorium in the country. For the 2026-2027 school year, generative AI will be off-limits for roughly 600,000 students from 2-K through eighth grade, companion chatbots will be blocked in every grade, and new screen time caps will apply systemwide. City leaders call it putting students first. We agree that tech use in schools must be purpose built for students and learning. But we are concerned that a blanket ban doesn't put students first and sidesteps the harder-but more useful-work of setting age-appropriate rules, safeguards, and adult oversight.

We agree with much of what is driving the policy. Unchecked access to AI poses risks for younger students and strong protections are needed. Children learn from human connection and teachers who know them, and by working through hard problems on their own. No tool should replace that. Schools need clear guardrails, and only tools with real evidence of impact belong in the classroom. We have said this for years.

A ban, though, is not a guardrail. It avoids the problem instead of managing it. A 14-year-old in New York can ask an AI chatbot for help the moment they leave the school building, and nearly everywhere else outside of it. Nearly nine in ten young people already use AI. The one place the city removed it is the one place where a trained teacher could sit with a student, show them how to use it well, and point out where it gets things wrong. This is also the one place where there are guidelines in place to ensure that the tools are purpose built to help learners and advance learning objectives. That is exactly where it should stay.

New York's economy shows why. Last week a CBRE report found that New York now has more tech jobs than the Bay Area. These students will graduate into jobs that expect them to understand AI, and banning it in school will not keep it out of their lives.

The policy also treats all technology as the same, and screen time is the wrong way to measure learning. Time on a device tells you how long it was open, not what a student was doing. A student scrolling a feed is not doing what a student does when annotating a primary source or using a reading tool set to their level. Education technology built for the classroom should not be judged the way commercial apps are. What counts is what a tool is for and whether it helps students learn, not how long a screen is on.

Reading shows what a blunt cap puts at risk. Many students now read through school and public library apps, where they can borrow more titles than a shelf can hold, hear audiobooks, enlarge type, and translate as they go. The library app Sora alone reported more than 2.76 million student checkouts in New York City schools in 2024-25. With the city's reading scores falling, schools should widen access to books, not cut off how students find and read them.

Parts of this plan are worth keeping, including high school pilots, the AI literacy classes, and the exceptions for assistive technology and multilingual learners. But accessibility should be more than an exception. Screen readers, captions, text-to-speech, and translation are how many students read and write at all. Treating digital access as something only some students get can single out the very students it is meant to help.

There is also a cost to getting this wrong. Districts have spent heavily on devices, curricula, and platforms, and a blunt limit strands the investment those tools were meant to support. A better path is managed access, with clear rules for how and when technology is used, so class time differs from lunch and younger grades from high school.

The mayor said the tech industry wants the city to believe AI-powered early education is inevitable and necessary. That is not our argument. We are not asking the city to hand young children to chatbots, only to keep the door open to careful, teacher-led use.

We have not seen the standards the city will use to review its tools, and we would like to. As the Technology in Schools Coalition begins its work, teachers should be at its center and the companies that build these tools are ready to help. Students deserve a policy that prepares them for what is coming, not one that pretends AI, or the tools students already learn with, can be kept out of the classroom.