Sometimes, and it depends on where you live. In most of the US and the UK, no law gives parents a general right to opt a child out of AI at school; schools choose tools the way they choose textbooks. But you can nearly always find out what is switched on, see your child's records, and ask for an alternative to one named tool. That narrow request usually works best.
Do schools need my permission to use AI with my child?
Usually not. In the US, the Federal Trade Commission's guidance on COPPA, the federal children's online privacy law, says a school can consent on a parent's behalf to an online service collecting a child's information, "but only if such information is used for a school-authorized educational purpose and for no other commercial purpose" (FTC, April 9, 2020). That is the mechanism that lets a district roll out a reading app, a math platform or an AI tutor to thirty children without thirty signatures.
The same guidance says the school "should give parents a notice of the websites and online services whose collection they have consented to." Should, not must. So the first thing you are entitled to ask for is not a veto. It is a list.
You also have a right that most parents never use. FERPA, the federal student records law, requires a school to let you inspect and review your child's education records "not more than 45 calendar days following receipt of a request," according to the US Department of Education's parent guide.
In the UK, the picture is similar: the school decides what is taught and with what. Under data protection law, parents can ask for the personal data a school holds about their child, and the Information Commissioner's Office explains how.
What has changed since August?
Three things happened in six weeks, and each one moved the answer.
Google switched the default. On August 10, 2026, Gemini in Google Classroom began reaching K-12 students of all ages. Google's admin notice of August 4, 2026 says it is "on by default for teachers and students of all ages" and is controlled by a setting in the school's admin console. Whether your nine year old sees a chatbot inside Google Classroom now depends on a setting somebody in the district office may or may not have touched. The notice does not mention parents.
Florida wrote parents into a rule. On September 16, 2026, Florida's State Board of Education is due to vote on an amended Rule 6A-1.0957, listed on the board's September agenda. As proposed, every district and charter school would have to adopt, by July 1, 2027, a policy that tells parents when an AI tool is approved: its name, the grades and subjects, how children interact with it and whether it gives automated feedback, and how to object. Parents would get a choice about their child using the tool, with information on a non-AI alternative "substantively comparable in instructional quality and access." Tools could not be designed to "simulate friendship, companionship, or an emotional relationship" with a student, and would have to keep records of student interactions for at least 30 days so parents can see them. Coverage has described the choice both as an opt-in and an opt-out; the proposed text speaks of parents opting their child into using the tool. Until the board votes, it is a proposal.
Parents went to court. On August 26, 2026, seven families sued Lower Merion School District near Philadelphia, arguing that a school cannot require children to use technology without their parents' consent. The case is about school devices and ed tech in general rather than AI in particular, and the district denies the claims. In Education Week's report of September 8, 2026, the lawyers quoted disagreed about whether the argument will succeed. It is a live case.
Add New York's ban on student use of generative AI from pre-K through 8th grade, which I read for families here, and you have schools pulling in opposite directions in the same month. Parents are caught in the middle, and many do not know where their own school stands. When the MassINC Polling Group surveyed 1,351 Massachusetts parents of K-12 students for EdTrust in Massachusetts (fieldwork October 21 to November 12, 2025, published January 2026), 37% were unsure whether their child's school had an AI policy at all.
What can I actually ask for?
You can ask for three things, in this order: the list, the records, and an alternative for one tool.
Start with the question I would ask first:
The Where Question is the one question to ask a school about AI: where does AI touch my child's work, and who checks it? It works because it cannot be answered with a policy. It has to be answered with a list.
The records matter if a tool gives your child feedback or keeps their chats: ask what it stores, for how long, and whether you can see it. Florida's proposal would require 30 days of interaction records for exactly that reason.
Once you have the list, decide which single item actually worries you. Then use what I call the Narrow Ask:
The Narrow Ask is how I would word any request to opt a child out of AI at school: one named tool, one subject, one alternative you can suggest. A narrow ask is a decision a teacher can make this week. A request to remove your child from "AI" is a policy question, and policy questions go to committees.
You could say, in an email to the class teacher:
"Could you tell me which AI tools my daughter uses in your class, and what she does with them? For the chatbot in her writing assignments, I would prefer she drafted on paper or in a plain document this term. I'm happy to talk about it, and I'm not asking you to change anything for the rest of the class."
That email gives the teacher a yes they can say without asking anyone. It also tells them you are an ally, which matters more than it should.
What I See in Schools
The pattern I see most often, working on AI strategy with schools, trusts and districts, is that the blanket letter and the narrow email land on different desks. A letter that says "I do not consent to my child using AI" goes up to the principal, who has to work out whether it covers the spell checker, the adaptive math app, the library search and the chatbot. Nobody can answer that quickly, so the answer is usually slow and vague.
A request about one tool in one subject usually stays with the teacher. Most teachers I train would rather adjust one child's assignment than defend a platform they did not choose.
The other thing I see: most schools have never been asked. Their policy was written for students and staff. The part that tells parents what is switched on is often missing, not because anyone is hiding it, but because nobody thought a parent would want to know. Your question is often the thing that makes them write it.
What will opting out not do?
It will not keep AI away from your child. Common Sense Media's 2026 census of 1,204 US children aged 9 to 17 (fieldwork March 2026) found that 44% of those who use AI reach it on a school computer or tablet (Common Sense Media, 2026). The same survey found children also reach it on their own devices and on shared family ones, inside apps that were never on any school list.
Opting your child out of the school's AI does not opt them out of AI. It moves their first meetings with it somewhere with no teacher in the room.
An opt-out also rarely covers the AI features built into tools the school already uses: the writing suggestions, the reading app that adjusts its questions, the search summaries. And it cannot teach your child the skill that actually protects them, which is checking a confident answer. That still has to happen somewhere. If it is not at school, it is at your kitchen table.
None of that is a reason not to ask. It is a reason to ask precisely, and to decide what you are protecting. My test for families is the same one I use with teachers: let AI help with the doing; keep the thinking with your child. If a tool is doing your child's thinking in a subject, that is the tool to ask about. If it is doing the admin, it probably is not.
How does this change with age?
Under 8. This is where the case for a narrow opt-out is strongest. New York's ban and Florida's proposal both treat the youngest children as a special case, and Florida's would add an extra age-appropriateness review for tools used from pre-K to 5th grade. Ask specifically about anything your child can type or talk to. An adaptive phonics app is not the same thing as an open chatbot.
8 to 12. Ask the Where Question, then pick your one tool. At this age the risk is less what the tool says and more what it does instead of your child: the first draft, the plan, the sum. The 9 to 12 guide covers what else they are meeting at home.
13 to 15. Bring your teenager into it before you email anyone. A 14 year old who is quietly removed from the class tool will often use a less careful one on their phone. You could say: "Show me what you use it for in class. If it's helping you think, fine. If it's doing the thinking, let's work out together what to ask the school for."
16 to 18. Treat it as their decision, made with you. They will be using these tools long after they leave, and the school may be the only place anyone teaches them to use them well. The useful request here is usually the opposite of an opt-out: ask how the school teaches students to use AI honestly and check what it says.
For a younger child who has been given the alternative, you could say: "I've asked your teacher if you can do your writing on paper for now. You haven't done anything wrong. I just want you to do that part with your own head for a while."
When does the school's decision come first?
In most places, by default: the school sets the curriculum and chooses the tools, and state or national rules sit above both of you. Four situations deserve care:
- If the tool is part of how work is graded or tested, an alternative may not exist. Ask how your child will be assessed fairly instead, rather than pressing for removal.
- If your child has an IEP or a 504 plan (in the UK, an education, health and care plan, or special educational needs support), raise it through that plan's review. A tool that reads text aloud or supports writing may be part of an accommodation, and changing it changes the plan.
- If something has already gone wrong, such as a tool saying something harmful or your child being upset by it, that is a safeguarding matter, not an opt-out. Tell the school's safeguarding lead or school counselor directly, in writing, the same day.
- If you are in the UK, remember that England's Department for Education sets expectations for suppliers rather than giving parents a veto. Its generative AI product safety standards, updated January 19, 2026, expect tools to avoid "names, descriptions, avatars or characters which could give an impression of personhood" outside short teaching tasks. They are voluntary, and they say nothing about parental consent, which is why the conversation with the school matters.
What can I do tonight?
Write down the one AI tool you have heard your child mention from school, and send the teacher the short email above. If you do not know of one, ask the school for its list. The Gemini guide explains what to check if your school uses Google Classroom, and how AI is used in schools walks through the four places it usually sits.
The next step
If your child's school has not yet told parents which AI tools it uses and how to raise a concern, that is part of the AI strategy work I do with schools: deciding what gets switched on, and making sure families are told. It is worth sending to your principal or head of year.
Florida's vote, the Lower Merion case and Google's defaults are all still moving. My Sunday newsletter tracks what shifted each week for more than 44,000 teachers and parents, without the panic.
Dan Fitzpatrick is a former secondary teacher and assistant headteacher, a Forbes contributor, and the founder of The AI Educator. Read Dan's story.
