You can block a named AI app on your child's iPhone in about two minutes, as long as that phone sits in your Family Sharing group. What you cannot block is AI itself. It is already inside Siri, Safari, Google and Snapchat, and none of that arrives as an app you can switch off. So the block is worth doing. Just do not mistake it for a finish line.
The two-minute version: blocking one AI app tonight
On your own phone, open Settings, tap Screen Time, tap Family, tap your child's name, then tap Apps & Websites. Find the app in the approved list, tap it, and tap Block App. You will need your Screen Time passcode, which is the four-digit code you set when you turned Screen Time on, and which should not be a code your child has watched you type.
Apple's own guidance, last updated on September 14, 2026, notes something worth knowing before you do it: a blocked app "appears dimmed on the Home Screen, but not deleted." Your child will see it there, grayed out, and they will know. I think that is a feature rather than a problem, but only if you are the one who tells them first.
You could say: "I've turned that one off on your phone for now. I'm not angry and you're not in trouble. I want to understand what you use it for before I decide what the rule is."
That sentence does more work than the setting does.
Why you cannot block "AI" the way you block an app
Here is the thing most guides leave out. "AI apps" sounds like a category, the way "games" is a category. It is not. On a modern iPhone, most of the AI a child meets was never installed by anybody.
It is in Siri and Apple's own on-device features. It is in Safari, in the AI answers that now sit at the top of a Google search. It is in Snapchat's My AI, which appeared one day at the top of a chat list in an app your child already had. It is in the games and the social apps, writing captions and rephrasing messages. Apple's parental controls do include a section for restricting Siri and Apple's AI features, and it is worth switching off if your child is young. But you will not find a toggle labeled "AI," because there is no such thing to toggle.
The scale of this is easy to underestimate. Common Sense Media's census of AI use by tweens and teens, published on June 8, 2026 and based on a survey of 1,204 children aged 9 to 17 in the United States, found 86 percent using AI and close to a quarter using it daily. The same study found that more than four in ten of the children surveyed said no parent or guardian had ever talked to them about using AI safely. Those two numbers sit badly together, and no setting reconciles them.
The Still-Allowed Check is what to do in the ten minutes after you block an AI app: open the apps you did not block, and find the AI inside each one. Blocking removes the app your child told you about. The check finds the three they did not, because those arrived as a feature inside something they already had, and nobody had to install anything.
Give it ten minutes on the sofa with your child's phone in your hand and your child next to you, not in another room. Open each app they use most. Look for a sparkle icon, a chat entry at the top of a list, a "summarize" or "rewrite" button. You will find more than you expect, and you will find it faster with them than without them.
The setting that stops the next one arriving
Blocking one app is retrospective. Age ratings are the setting that works ahead of you.
In the same place, under content restrictions, Apple lets you apply age ratings to apps so that anything above the band you choose cannot be downloaded at all. Apple rebuilt these bands in July 2025, replacing the old set with 4+, 9+, 13+, 16+ and 18+, and the new bands show on devices running iOS 26 and later. Look up the rating the AI app your child uses actually carries, then set the band below it. The app then stops being downloadable, rather than being blocked after the fact.
Apple also turns Ask to Buy on by default for children under 13, which means a download request arrives as a message on your phone for you to approve or decline. If your child is under 13 and you have never seen one of these requests, that is worth checking tonight.
None of this substitutes for the age rule the company itself sets. OpenAI's help center states plainly that ChatGPT "is not meant for children under 13" and that those aged 13 to 18 need parental consent before using it. But as I have written about the AI age limits before, the age limit is a promise, not a lock. It is a date of birth typed into a box, usually by the child, and nothing about the phone enforces it.
What to do when the phone is not in your family group
This is where a lot of parents get stuck, and the honest answer depends on your child's age.
Apple's rules, as of September 14, 2026, are that a child under 13 has to be part of a Family Sharing group. Between 13 and 17, if you have set up Screen Time, your child needs permission from the family organizer to leave it. At 18 they can leave, and manage their own settings, and that is that.
So if your fifteen-year-old's phone is in your group, you have the controls described above. If it is not, and they set it up on their own account, you do not have a technical route. You have a conversation. Getting them to join a family group at fifteen, having been outside it, is a negotiation, and it goes better as a trade than as an instruction.
It is also worth knowing what these controls never gave you. Apple tells children this directly: a parent managing their contacts and communication still "can't read your messages or listen to your conversations." Screen Time decides what your child can open. It was never built to show you what they typed, and AI chats are no exception, which is a subject I have written about separately in what you can and cannot see of your child's ChatGPT history.
The mistake I see parents make
I write about these tools for Forbes, and a great deal of what I end up writing about is AI being switched on somewhere nobody announced it. That is the part that makes this question harder than it looks.
The mistake is treating the block as the end of the job. A parent spends twenty minutes on Screen Time, gets the app grayed out, feels the relief of having done something, and stops. Two weeks later the child is using the same model through a browser, or through the assistant inside a game, and nothing in the house has changed except that the parent now believes it is handled. The pattern I see most often is not a child defeating a control. It is a control that was never covering the thing the parent was worried about.
Blocking an AI app removes the one your child asked permission for. It leaves the ones that arrived without asking.
And there is a second gap, which is the school's device. Whatever you set at home stops at your front door. A school laptop or tablet runs on the school's account, with the school's filtering, and your Screen Time passcode means nothing on it. I have written about what a school can actually see of a child's AI use, and the short version is that the answer is different on every device your child touches. If your child's school has never written to parents about what its devices allow and what its staff have been trained to do, that gap is worth naming out loud, and it is the work I do with schools through an AI readiness review.
How this changes by age
Under 8. Block generously and explain briefly. At this age the AI they are most likely to meet is a voice assistant answering a question they were about to wonder about. You could say: "Let's guess first, then we'll ask."
8 to 12. These are the years to set the age ratings and turn Ask to Buy on, because this is when apps start arriving through friends rather than through you. Do the Still-Allowed Check together and make it ordinary rather than an inspection. You could say: "Show me the ones with the little sparkle on them. What do they do?"
13 to 15. Blocking alone starts to fail here, because this is the age of the second account and the logged-out browser. Aim for a phone that is in your family group with fewer restrictions rather than one outside it with none. You could say: "I'd rather know what you're using than police it. What's the one you'd be annoyed to lose?"
16 to 18. Controls are mostly the wrong tool now. What is useful is the habit of checking a machine's answer, because they are about to take it into exams, applications and a first job. You could say: "What's the last thing it got wrong?"
What blocking cannot guarantee
Being straight about the limits is the only way any of this is useful.
A block on the app does not block the website, unless you also restrict web content. It does not cover a friend's phone, a shared tablet or a games console. It does not cover the school's device, which runs under rules you did not set. It does not stop a second account being made on a different email address. And every parental control on every platform can be worked around by a determined teenager with time and a search engine, which is not a reason to skip them, only a reason not to rely on them alone.
Two things take precedence over anything in this article. Your child's school sets the rules for its own devices and its own work, and where those conflict with your home rule, the school's rule applies at school. And if what is worrying you is not the app but your child, if something they have said or written has frightened you, that is a conversation for your doctor, your pediatrician or your school's safeguarding lead, not a settings menu.
What I would do this week
Block the one app you know about. Then spend ten minutes doing the Still-Allowed Check with your child beside you, because the apps you did not block are where the interesting answer is. Set the age ratings so the next one needs your approval. Then ask your child's school the one question a settings menu cannot answer: what does AI do on the device you lend my child, and who checks it?
If you want the next of these as it is written, I send one email a week to parents on exactly this, and you can sign up here.
Dan Fitzpatrick is a former teacher and assistant headteacher who now works with schools on AI. More about Dan.
