Yes, if an adult teaches them how, and your school's policy is not what decides that. A ban and a rollout can each hand back a child who uses AI well, or one who uses it badly. What separates them is whether anyone sat with your child while they judged an AI answer. Follow your school's rule either way: it covers the school day, not the phone in their bag.
Why are schools banning AI right now?
Because the adults who work in them lost confidence faster than the tools improved. Three years of quiet adoption ended in a few weeks this fall.
New York City paused student-facing generative AI from pre-kindergarten through 8th grade for a year, about 600,000 children. I wrote about what the New York ban does and does not cover when it landed. Los Angeles went further and quieter. On September 2 and 3, 2026, according to LAist's reporting, the district switched on web filtering that blocks any site categorized as an AI tool across its student Chromebooks and tablets, reaching about 378,000 students. Until then, students aged 13 and over had been allowed to use AI under supervision once they had completed a digital citizenship lesson. Board member Nick Melvoin, looking at the room at the committee meeting where it was announced, said: "Just based on the facial expressions of folks in this room, many parents, some of whom work at the district, some of whom don't, I don't think it's well-publicized."
Hold that detail, because it is the whole argument in miniature. Los Angeles did not move from permission to prohibition. It moved from supervised and taught to blocked, and the parents found out afterward.
Axios reported on September 7, 2026 that candidates for the Chicago school board have pledged a three-year moratorium, and quoted parents describing today's students as "the guinea pig generation."
Teachers are not the holdouts here. They are the ones asking. In the NPR/Ipsos poll of 545 US K-12 teachers fielded April 27 to May 5, 2026, with a margin of error of five points, 55 percent said AI is mostly a shortcut for students to avoid doing work, 54 percent said it is making critical thinking harder to learn, and 59 percent said it is eroding trust between students and teachers. Sixty-two percent of those same teachers use AI for their own work.
That is the mood a ban comes out of. It is a reasonable mood. It is not, on its own, a plan.
Does banning AI at school actually work?
It works on school devices, for six hours, and nowhere else. That is not a cheap shot at the policy. It is the policy's own scope.
A filter on a district Chromebook does not reach a phone in a bag, a laptop at home, a family tablet, or the AI summary sitting at the top of a search results page. Nor does it reach homework, which is where most of the friction lives. So a ban does one thing well, which is to take AI out of the lesson, and one thing badly, which is to decide anything at all about the six hours after the last bell.
There is a second cost, and it is the one I would put to a school leader. A pause that nobody explains teaches children that AI is contraband. To a fourteen year old, contraband is a recommendation.
None of which makes a ban wrong. For the youngest children I think it is right, and I have said so. A seven year old learning to read, to form letters, and to hold a number in their head gains nothing from a chatbot that a teacher and a book cannot give them better. A pause with an end date and a review is a decision. A school with no policy at all is not.
So should I want my child's school to allow it?
Not as such. Wanting the policy is wanting the wrong thing.
The evidence points at a different question. RAND surveyed nationally representative panels of US teachers, school leaders, district leaders, students and parents in spring 2025 and published the results on September 30, 2025. Fifty-four percent of students and 53 percent of English language arts, math and science teachers said they used AI for school, both up by more than fifteen percentage points on the year before. Only 35 percent of district leaders said they provided students with any training on AI. And more than 80 percent of students said their teachers had never explicitly taught them how to use it.
Set that next to what parents say they want. Child Trends published an analysis on August 25, 2026 of 720 US parents and caregivers of children aged 12 to 18, surveyed through NORC's AmeriSpeak panel in December 2025. Adding up the answers, roughly six parents in ten wanted responsible AI use taught across subjects and another three in ten wanted it taught in a specific class. Only 13.5 percent wanted it not taught at all. On outcomes the same parents split almost evenly: 34 percent thought AI would help their child's learning, 36 percent thought it would harm it.
So: nearly nine parents in ten want it taught. More than eight students in ten say nobody taught them. Both figures are from the United States, both are recent, and neither is about banning anything. The argument in the news is about the wrong variable.
This is what I mean by the second policy.
Every school has two AI policies. The written one, which is on the website, and the real one, which is whatever happens when thirty children open a laptop and one adult is in the room. Parents read the first. Children live in the second. A ban and a rollout can each come with a good second policy or with none at all, and that, not the ban, is what decides what your child learns.
A ban and a rollout can hand back exactly the same child: one who uses AI most days and has never once been asked to say why it was wrong.
The good version of a ban is a school that switches the tools off, tells the children why in language they can repeat, protects handwriting and mental arithmetic and first drafts, and says out loud what happens at home. The bad version is a filter and a silence.
The good version of a rollout is a school where children are regularly handed something an AI produced and asked to find what is wrong with it. The bad version is a license and a silence. The silence is the thing the two bad versions have in common, and it is what your child brings home.
What I See in Schools
I train teachers on this, and the same few patterns turn up whichever way a school has jumped.
When a ban lands, it lands unevenly. The filter goes on at the district level over a weekend. On Monday the science department, which had been using an AI tool to generate practice questions, finds it gone and quietly goes back to the photocopied worksheet. The English department, which had built a whole unit around critiquing machine-written paragraphs, loses the unit. Nobody planned either of those things. Meanwhile the oldest students, who mostly work on their own devices anyway, notice nothing at all. Within two weeks the staff room has divided into people who think the ban is a relief and people who think it is an embarrassment, and no child has been told anything beyond "it doesn't work anymore."
When a rollout lands, it lands on the enthusiasts. A senior leader announces the tool in a twilight session in September. Perhaps a fifth of the staff try it that term, and those are the ones who were already interested. The training is one hour, and it is about how to log in and what not to put into it, because that is what the safeguarding paperwork requires. It is almost never about how to teach a child to argue with the output, because nobody has been trained in that either. The policy document says pupils will be taught to use AI critically. Ask which lesson, in which subject, in which term, and the answer is a pause.
Two more things I notice, and parents are often surprised by both.
The first is that the most common AI conversation in a school is not about cheating. It is about a piece of work that is correct, tidy, well organized and completely silent about the child who handed it in. No wrong turn, no odd example, no sentence only that student would write. Teachers describe this before they describe plagiarism, and they describe it in almost the same words in very different schools.
The second is that the school's AI policy is frequently newer than the teacher's knowledge of it. I have stood in front of rooms in the same week that a district announcement went out and found that most of the people in the room had not seen it. This is not negligence. It is what happens when a technology decision is made above the level of the classroom and communicated by email in September. It is also why a question to the school that can only be answered with a policy will get you a policy, and tell you nothing.
So ask a question the written policy cannot answer.
What to ask the school
Three questions. None of them is "do you allow AI?", because that has only a policy for an answer.
- The Judging Question: how often is my child asked, in a lesson, to judge something an AI produced and say why it is wrong? This is the one classroom habit the international data links to higher scores among children who use AI every day. The evidence behind it is in what the research actually says about AI and children's learning.
- If AI is switched off here, what have you told the children about why, and what are you teaching them about it while it is off? A good answer names a lesson or an assembly. A weak one names a filter.
- Where does work still get done without any help, and how often? A child who writes in class regularly cannot quietly stop practicing, whatever the policy is.
A good answer to any of these has a subject, a term and a person in it. "We are following the guidance" is not an answer, and you are allowed to say so pleasantly and ask again.
If your school allows AI and you would rather your own child did not use a particular tool, that is a narrower and much more winnable request than it sounds: what you can actually ask for when opting out sets out how to word it.
Is it different in the UK?
Yes, and almost in the opposite direction. England has no national ban, and the Department for Education and the Department for Science, Innovation and Technology have been pushing the other way, hosting an international AI in Education summit in January 2026, publishing updated generative AI product safety standards, and working on curriculum-aligned tutoring tools aimed at supporting up to 450,000 disadvantaged pupils. The Schools White Paper published early in 2026 states that "teachers remain at the heart of classroom practice." Whether an individual primary or secondary school switches AI on or off is still that school's own decision, so the questions above are the ones to take to a parents' evening, or a parent-teacher conference if you are reading this in the United States.
The gap is the same on both sides of the Atlantic. One government is blocking and one is funding, and neither has yet made sure a child is taught to doubt what comes out.
What to do at home this week, by age
The point of all of this is that the gap is fillable, and filling it takes about five minutes a week.
Under 8. Nothing to fill yet. Match a school ban at home without discussion, and if your child meets an AI answer through a smart speaker or a search result, do the noticing out loud. You could say: "That computer is good at copying and not good at knowing. Let's see if it got it right."
8 to 12. This is the search-summary age, and the habit is cheap to build. Once this week, read the AI answer at the top of a search with them and find one thing to check. Let AI help with the doing; keep the thinking with your child. You could say: "Read me the bit at the top. Now, how could we find out if that's true?"
13 to 15. Assume they use it whatever the school decided, and ask about the work rather than the tool. You could say: "I'm not going to ban it. Show me the last thing it got wrong for you." That question does more than a rule, because it asks them to have noticed.
16 to 18. Exams and applications are the real stakes now, and the school's rules and the exam board's rules both outrank yours. You could say: "If this tool disappeared the week before your exam, what would you be shaky on? Let's start there."
If you want the longer version of the home habit, teaching a child to question an AI answer is the practical one, and what your child's teachers are actually doing with AI is worth reading before any conversation with the school.
What no school policy can guarantee
No policy, filter or setting can tell you that your child thought. A ban cannot stop them using AI at home, a rollout cannot make them learn from it, and a filter on a Chromebook has no view of a phone. The only reliable check is your child explaining their own work to you for two minutes, which is why that beats any control panel.
Where the school's rule and this article disagree, follow the school. If your child has an education plan that includes assistive technology, whether a 504 plan, an IEP or an EHCP, that plan comes first and the tools in it are there for a reason. And if what is worrying you underneath this is your child's learning rather than AI, an AI rule will not fix it; that is a conversation with the class teacher, and sometimes a specialist, and it should not wait for a technology debate to finish.
The reason to feel steadier at the end of this is not that the bans are wrong or that the rollouts are right. It is that the thing that decides the outcome turns out to be small, local and available to you tonight.
If your child's school is somewhere between switching AI off and switching it on and has not yet said which, finding out where a school actually stands is the work I do with schools, and it is worth forwarding to a head of year or a principal. I also write a newsletter that keeps teachers and parents up to date on this without the panic, and parents are very welcome.
Dan Fitzpatrick has trained more than 150,000 educators in more than 30 countries, and is the founder of The AI Educator. More about Dan.
