New York schools block 38 AI tools after safety review

New York City's new classroom AI policy has been widely described as a ban on ChatGPT — but the more telling detail may be the fate of three dozen other programs quietly losing their AI features.

The city's education department will remove or disable artificial-intelligence functions in more than 38 software products that were previously approved for school use but now fail the district's updated safety and oversight standards.

The move came as Mayor Zohran Mamdani and Schools Chancellor Kamar Samuels unveiled what they called the city's first comprehensive AI policy for public schools, a one-year framework covering close to 600,000 students — roughly two-thirds of the city's public-school population, according to CNN.

Under the new rules, generative AI is off-limits for students in 2-K through eighth grade. That means no ChatGPT, no Claude, and no AI-powered companion chatbots at any grade level.

High school students will get access only to a limited set of approved tools, with instruction centered on what the district calls "AI critical thinking."

The 38-plus software programs being retroactively blocked suggest the review is not just about the big-name chatbots. These were tools that had already cleared the district's approval process.

Their removal marks a sudden reversal of status for products that schools had been using, and it signals a new willingness by the district to treat earlier approvals as provisional.

Mamdani framed the policy around human connection and child development. "Children need their teachers and human connection in order to learn and in order to grow," he said at the announcement, as reported by PTI.

He also pushed back on industry pressure, arguing: "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."

The mayor went further, questioning whether independent evidence actually shows AI benefits for elementary and middle-school students, and pointing to the risk of research funded by companies that profit from the technology.

That skepticism appears baked into the new safety review: vendors who once won approval may now need to prove their tools meet stricter standards around safety, oversight, and educational value.

The restrictions are not absolute. Students with disabilities and English-language learners can receive exceptions so their education is not disrupted.

Teachers may still use AI for planning lessons and operational tasks, but they are barred from using it to grade or assess student work.

The city is also introducing grade-based limits on individual screen time.

What happens next is deliberately open-ended.

The policy runs for one year, and a new Technology in Schools Coalition — made up of educators, parents, students, and other stakeholders — will study the results and advise the city on future AI regulation.

That makes this a pilot of sorts, not a permanent settlement.

Our analysis suggests the 38-tool purge carries a broader warning for the education technology industry. School contracts have long been pursued as stable, recurring revenue.

This review demonstrates that a single policy shift can instantly turn approved products into non-compliant ones, creating a new kind of regulatory risk for vendors.

The precedent set in New York — the largest district in the country — may also influence other districts weighing similar oversight measures, meaning edtech companies could face a patchwork of stricter reviews nationwide.

The reversal also fits a longer pattern of whiplash in New York. The state separately banned personal smartphones during the school day last year.

The education department blocked ChatGPT on school devices and networks in 2023, then reversed that decision months later.

The latest policy is a renewed attempt to draw firmer lines, but with a one-year sunset clause, it leaves room for another reversal.

For families, the immediate steps are straightforward: ask your school which of the 38 affected programs were in use and what alternatives will replace them; review the district's screen-time limits; and talk to children about why ChatGPT and Claude are no longer available in class.

For small edtech companies, the lesson is more structural: build AI features that can be switched off without destroying the product's core value, document safety and data practices clearly, and avoid relying on a single district's approval as a long-term guarantee.

The New York review is a reminder that in education, trust from regulators can be granted quickly — and withdrawn just as fast.

Credits: This article draws on reporting by CNN and PTI, as reflected in the provided source material.