The good, bad and ugly in the new AI/Screentime guidance

Sept. 10, 2026

Welcome back to a new school year! I hope you all had a great summer and a good Labor day break. There is a lot happening.

Last week, the Mayor and the Chancellor announced new AI/screen time policies to be implemented this school year, while longer term guidance will be developed.  Some elements of these policies are a decided step forward from the unrestrained AI/ed tech assault that students in NYC and elsewhere have suffered over the last few years – but much of what these policies include is ambiguous or inadequate to deal with the crisis our schools are facing.

Here is the good, bad and ugly in this new guidance.

First the good:
For grades 2K-8: All student-facing generative AI will be prohibited in grades 2K-8, including AI tutoring, chatbots, instructional programs, and intervention software. That means no Google Gemini, and no Amira, unless the AI portion of the software can be “turned off”, which seems unlikely. In fact, the DOE claims to have already disabled nearly 40 AI products.

Yet a sticking point is how DOE will confirm that the AI component has either been removed or never embedded in the first place, even as it has been rapidly infused into hundreds of ed tech programs over the last two years. Already, teachers say their schools are being spammed with vendors promising to turn off the AI in their products, but how will this be verified?

In the past, the DOE has admitted that they have no idea which ed tech products use AI, and those in charge have done a very poor job overseeing the use of ed tech in general, even when it comes to ensuring it complies with the state student privacy law, Ed Law 2D, passed in 2014.

One tech expert told us that it is difficult if not impossible to monitor whether ed tech programs are using AI, unless the software is downloaded and operated on DOE’s own servers, which so far, they have refused to do. And what about the Google search engine which now runs with AI; will students be told not to use Google and instead a different privacy-protective/non-AI search engine, like DuckDuckGo?

Now the bad:
In any case, AI tools will still be assigned to all high school students in mandated classes on “AI literacy,” those enrolled in relevant CTE/vocational programs, plus students who attend up to 180 unspecified high schools in “pilot programs” enrolling up to 50,000 students.

In these 180 schools, one of five approved AI programs used will be either Quill, Edia, Brisk Teaching, or PlayLab, with more details about these particular programs available in the DOE official press release here.

Another big loophole in the new AI guidance is that AI products can still be assigned to students in any grade if they have disabilities or are English Language Learners, as long as the DOE considers these programs “necessary.” Will parents be told which AI programs their children are assigned to use, why they are deemed necessary, and allow them to consent or opt out, as they are able to do with IEPs? The guidance is so far silent on this.

Teacher use of AI will remain mostly unrestricted, except for grading and high-stakes decision making, but even what the DOE means by these terms remain unclear. Can teachers use AI to correct papers, for lesson planning, or to provide individual student support or communication? And if so, is that acceptable considering so many of these products have been shown to generate misinformation, hallucinations, and even racial bias?

While it says that “all special education documents are developed by qualified professionals” does that mean these professionals can use AI to write IEPs, and if so, what protections will be required to ensure that the sensitive, personal information of these students does not leak out in ways that could risk their privacy, health, or safety?

Finally, DOE has said nothing about whether students will be prohibited or even discouraged from using off-the-shelf AI tools to do their homework outside of school, even though many of these products are inherently unreliable, data-mine personal data in ways that violate state law, and undermine student critical thinking, creativity and mental health. Shouldn’t there be something in the policy to address these concerns?

The ugly: Screen time guidance
If the new AI policy is full of loopholes, the new screen time guidance is even worse.

Students in grades 2K-2 are supposed not to use individual devices, yet screeners, assessments, and diagnostics are all exempted from these restrictions. The DOE has not specified which of the excessive number of ed tech products currently used are going to be defined this way, and already principals are telling parents that many of the programs their kids were assigned to use in the past will be described as such.

For students in 3rd-5th grades, there are recommended limits for devices of 30 minutes per day, and 45 minutes per day for grades 6-8. Yet again, these so-called screeners, assessments, and diagnostics as exempted, as well as any assistive technologies deemed “necessary” for SWDs and ELLs.

For high school students, there are no screen time limits at all, even suggested.

Moreover, any teacher-led use of computers, tablets, or interactive smart boards for whole or small group instruction are not to be counted (whatever that means), as well as “centrally approved instruction (e.g. e-books, robotics, coding, and simulations).”

Finally, there is nothing in the guidelines about what oversight will be exerted by DOE to encourage that schools stick to these recommendations, or whether parents will be told how much time their students are spending on screens. Already some principals have been told that they can evade even these loosely recommended limits by assigning two or more students to a device.

Next steps and how you can have your voice heard
The DOE has said that starting this month, a new “Technology in Schools Coalition” will begin to deliberate on permanent policies on AI and screentime that will be implemented next year.  This group is supposed to include students, educators, parent leaders, elected officials, advocates, union partners, and “experts in the field.”

Yet only member announced of this group so far is Julie Samuels, President of Tech: NYC, an industry trade group representing ed tech companies, including Google, Open AI and others  that sell products to our schools. Samuels has already written that she strongly opposes any AI ban in any grade.

We have asked the DOE a number of follow-up questions, including how people who are interested in having a voice on these issues can apply to be a member, along with other questions to try to clear up the many ambiguities that remain. You can see these questions here. If and when we get answers I will add them to the page. You are also welcome to ask some of these questions and add your own to your principal and superintendent.

In any case, on the very same day as the Mayor’s press release announced that these new policies were the “most expansive student-facing AI moratorium in nation,” the Los Angeles Superintendent announced a moratorium on the use of all AI products by students in all grades – without exceptions or exemptions.

Due to parent demand, LA schools had already instituted mandatory screen time limits in all grades, not just advisory, and said that the district would keep track of whether their screentime limits are adhered to via software installed on school-issued devices

Meanwhile, several important events are scheduled for next week.

·       On Monday, Sept. 14 at 7 PM, the AI Moratorium Coalition will be holding a Zoom call to discuss the new guidelines, hear from parents about what’s happening in your schools, and collaborate on strategies moving forward. You can register here.

·       The next day, Tuesday, Sept. 15, at 1 pm at City Hall, NYC Council Committee on Oversight & Investigation and Committee on Education will be holding joint hearings on Screen Time and the Use of Digital Devices in schools. You can sign up to speak either in person or remotely here. It would be great if you could show up in person, especially as a rally calling for strengthening the DOE’s screentime policies will be held at City Hall Park beforehand at 12 noon.

·       Finally, that same evening, Tuesday Sept. 15 at 7 PM, the FTC and the NTIA (National Telecommunications and Information Administration) will be holding a “Listening Session on Screen-Based Instruction in K-12 Schools”. More info and to sign up to speak or listen is here.

The FTC has shown a real interest lately in tightening up and/or enforcing existing federal privacy laws and is now considering trying to restrict excessive screen time in schools via their E-Rate program, which provides funding for internet services to schools. Each participant will be provided with three minutes to speak, which is more time than parents get at a PEP meeting. If you want talking points, please reply to this message.

We will also be holding a briefing session very soon on what rights parents have in state and federal law regarding their kids’ privacy and the use of ed tech, along with strategies you can use to opt out and fight back.

Hope to see you Monday night and/or Tuesday, and thanks!

Leonie Haimson
Executive Director
Class Size Matters
124 Waverly Pl.
New York, NY 10011
leonie@classsizematters.org