Last updated October 18, 2026: California's chatbot bill, which was waiting for the Governor's signature when this was first published, was signed on September 10. Los Angeles has gone further than New York, Chicago's board candidates have a pledge on the table, and several details of New York's policy are now stated as the city's own release and Chalkbeat give them.
New York City has paused student-facing generative AI for every public school student from pre-K through 8th grade for the 2026-27 school year. That changes what happens on a school device and leaves your evenings as they were, because homework, personal phones and the AI answers at the top of Google are not the school's to switch off. Your own school's rules, not New York's, apply to your child.
What did New York City actually decide?
On September 2, 2026 the Mayor's Office announced a one-year moratorium on student-facing generative AI from the earliest pre-K classes through 8th grade, which is roughly ages four to fourteen (Nursery to Year 9, if you are reading in the UK). Companion chatbots are banned on school devices in every grade. The release also sets screen-time limits: restricted one-to-one device time for 2nd grade and below, and recommended caps of 30 minutes a day in grades 3 to 5 and 45 minutes a day in grades 6 to 8.
High school is different. Up to 50,000 high school students can take part in pilots of five named tools, supervised by a trained educator, and every high schooler gets two 45-minute AI literacy modules a year. Teachers can keep using AI for planning and operational tasks if the tools meet the district's safety standards. Assistive technology for students with disabilities, support for multilingual learners and career programs such as computer science are exempt.
Chalkbeat's reporting that day adds detail the release does not. An Amira spokesperson said the reading tutor was in use in 222 schools, and a city adviser said Amira would turn its AI components off. Gemini comes pre-loaded on many students' Chromebooks; a city official did not say directly whether it would be disabled, though advocates told Chalkbeat they had been told it would be shut off. The release says a Technology in Schools Coalition will assess the policy through the 2026-27 school year and publish recommendations for future years. Chalkbeat reports the ban lasts "for at least this coming school year".
Is New York the only place doing this?
No, and the others are not the same policy. This is what each had put on record when I checked on October 18, 2026.
| Where | What was decided | Who it covers | Status |
|---|---|---|---|
| New York City | One-year pause on student-facing generative AI; companion chatbots banned in every grade | Pre-K through 8th grade; limited high school pilots | In force for 2026-27; coalition to publish recommendations |
| Los Angeles Unified | District restricted all student access to generative AI platforms | All students, all grades | In force for 2026-27; committee to report to the board by the end of the school year |
| Chicago | A pledge, not a policy: a three-year ban on AI for students and staff | 24 of about 45 school board candidates had signed by September 10 | Not in force; depends on who wins |
| Norway | Government said students in elementary schools may not use generative AI from the fall term | Elementary schools | Announced June 2026 |
Los Angeles matters if you live near it, because it is broader than New York. K-12 Dive reported on September 3 that the district had restricted all student access for the school year while it reviews instructional use and safeguards; before that, students 13 and older could use approved tools after completing digital citizenship instruction. SAN's coverage reports the ban applies to district-owned devices, includes Gemini and the chatbots inside Google's classroom software, and leaves the AI summaries in Google Search untouched. SAN also quotes the district's interim chief information officer saying a student's activity can be tracked if they use AI on a personal device while logged into a district account. Treat that as one official's description, not a published policy, and ask the school.
In Chicago, the pledge was crafted by Illinois Families for Public Schools, according to the Sun-Times, and Chalkbeat first reported it on September 1. It also promises no screen time before third grade. A candidate's signature is a promise, and I would not plan around it. In Norway, the government said in June that students in elementary schools would not be allowed generative AI tools from the fall term, with more money for books in place of tablets.
What does the ban not cover?
It does not cover your home. Chalkbeat notes that the new guidance does not address how to respond to students using AI outside of school. So the child who cannot open Gemini on a school Chromebook at two o'clock can still ask Google a question at four o'clock and get an AI-written answer at the top of the page.
That matters more than it sounds. Common Sense Media's census of 1,204 US children aged 9 to 17, surveyed in March 2026 and published on June 8, 2026, found that 86 percent use or interact with AI and 24 percent do so daily. The most widely used form was the AI-generated summary in search results, which 75 percent had used or interacted with. The Common Sense Media assessment of Google's AI Overviews and AI Mode, updated July 14, 2026, rates them "Unacceptable Risk" for kids and teens, and says "No one can turn the AI features off", neither a parent nor a school administrator nor the child.
The same census asked whether a parent or guardian had talked with the child about how to use AI safely. 44 percent of all the children surveyed said no. A school pause does nothing about that number. Only you can.
Is the ban a good idea?
For the youngest children, yes, and I say that as someone who has spent the last few years helping schools, trusts and districts write AI strategies, and who has advised the Department for Education, KHDA in Dubai and Kazakhstan's Ministry of Education on the same question. Learning to read, to write by hand and to hold numbers in your head happens best with a person, and a seven year old gains nothing from a chatbot that a good teacher and a book cannot give them. A pause with an end date and a review, which is what New York has chosen, is a decision rather than a mood. I would rather see that than a school with no policy at all, and I would rather see it than a permanent ban that nobody ever revisits.
Where I part company with the cheering is here. The school day is six hours. Most of what an eleven year old meets of AI is not in it. A pause that is never explained to children teaches them that AI is forbidden, and to a thirteen year old, forbidden is a recommendation. And the pause covers students, not the questions parents have been asking me all year: is my child using it, for what, and what should I say.
A school can decide what happens on a school device between the first bell and the last. Everything else gets decided at the kitchen table, whether or not anyone decides it.
I call this the Kitchen Table Question, and it is the one I ask parents whenever a school announces an AI policy: what does this change about what my child meets at home tonight? If the answer is nothing, the policy is the school's decision, and the rest is still yours. I wrote the version of this argument for school leaders in September, if you want to send your head something more formal. And if you are weighing the ban itself rather than this district's version of it, the part worth watching is what replaces it.
What I would do at home this week
Match the pause for younger children, and use it as an opening with older ones. By age:
Under 8. The school's pause is what I would do at home anyway: no chatbots, no AI toys that talk back, and any AI they meet, they meet with you. If a smart speaker or a search result answers them, that is a chance to say something simple. You could say: "Some computers can talk like people. They are clever at copying, not at knowing. If one ever says something that worries you, come and tell me."
8 to 12. This is the search-summary age. Sit with them for one search this week and read the AI answer at the top together, then check it against a real page. If their school has paused AI, mirror it for homework this term, and let AI help with the doing while you keep the thinking with your child. You could say: "Your school has pressed pause on AI this year so you can get really good at the things it can't do for you. At home we'll do the same for homework. If you want to try it, we try it together."
13 to 15. Assume they use it. Pew Research Center, surveying 1,458 US teens and their parents from September 25 to October 9, 2025, found that 64 percent of teens report using chatbots and that 54 percent of all US teens have used them to get help with schoolwork, while only 51 percent of parents say their own teen uses chatbots. Ask to see how they use it, agree which parts of a piece of work are theirs, and set up the parental controls that now exist on ChatGPT. You could say: "I'm not going to pretend you don't use it. Show me how you use it for school, and let's agree which bits are yours."
16 to 18. No pause reaches them, and the real questions are exams, coursework and applications. Ask them to teach you something with an AI tool this weekend. You will learn how they think, and they will hear themselves explain the limits.
No setting can promise that any of this holds. A school's pause governs school devices and accounts only, and where your school's rule, your state's law or your pediatrician's advice says something different, theirs comes first.
What to ask the school
Five questions, whether or not your school is in New York or Los Angeles:
- Which tools are being switched off, and which stay for children with additional needs? If your school uses Google Classroom, ask specifically whether Gemini is on, as Google switched it on by default for school accounts of every age in August wherever the school had already allowed it.
- What will you be looking at during the year to decide what happens next?
- What are you teaching my child about AI while it is paused? In New York, only high schoolers get the literacy modules.
- Does the restriction reach my child's personal device if they sign in with a school account?
- What do you want us to do at home about homework?
A good answer names tools, names a review date, and gives you something to do. A weak answer is "we are following the guidance".
What to watch next
New York's coalition and Los Angeles's committee both report by the end of this school year, and Chicago's pledge depends on an election. The law with the longest reach so far is in California. Senate Bill 1119, known as Adam's Law, was signed on September 10, 2026. Common Sense Media's summary says it requires companion chatbot operators to carry out and document an annual child safety risk assessment, sets default protections for child users including parental controls, notices and time limits, and requires protocols for suicidal ideation, sycophancy and isolation, with in-app crisis referrals. It gives an effective date of July 1, 2027. It is about companion chatbots, not school tools, so it does not change what your child's school allows; what the California laws mean for your family sets out the rest.
If your child's school is working out what to do about AI this year, this is the work I do with schools: a strategy that says what AI is for, what students may do, and what changes on Monday morning. It is worth sending to the head. And if you want to keep up without the panic, my Sunday newsletter goes to more than 44,000 teachers and parents each week.
Dan Fitzpatrick is a former teacher and assistant headteacher, a Forbes contributor, and the founder of The AI Educator. Read Dan's story.
