What Each Platform Lets You Switch Off
Checked in August 2026, and worth confirming yourself, because these change several times a year. OpenAI added parental controls to ChatGPT in late 2025. You link your own account to your teen's, and from there you can set quiet hours, turn off voice mode, turn off memory so the tool stops carrying context between chats, turn off image generation, and opt your teen out of having their conversations train future models. Linked teen accounts also carry extra content protections by default.
Google takes a different route with Gemini. A parent enables access for an under-13 through Family Link, and image generation stays blocked for anyone under 18. Supervised access is not offered in the EEA, Switzerland or the UK, and the rollout has been gradual, so it may not appear for you yet. Anthropic requires Claude users to be 18 with no parental consent path, which means there is no control layer to configure and the question does not arise.
If you want the ten minutes in order: create your own account first, link it to your child's, work down the settings list turning off anything your child does not need, then sit with them and open the tool once together so they see what changed and why. Doing it in front of them rather than overnight is the part most parents skip, and it is the part that decides whether the arrangement survives the month.
The Gap Nobody Mentions
Set all of that up and you will notice something missing. The controls give you switches, not transcripts. You can decide when your child uses the tool and which features they get, and you will learn close to nothing about what they discussed, because the main platforms offer no conversation access and no topic summaries.
That is worth sitting with rather than working around. The settings are a boundary, not a window, and a parent who sets them up and feels informed has been given the wrong impression. If you want to know what your child is using it for, the way to find out is the old way: ask them, look at the work, and stay in the room often enough that it is normal.
Working out the shape of the week is easier with a plan you can start from.
Get the GuideA simple step-by-step plan for getting started.
Controls Attach to Accounts, Not to Children
Every one of these settings applies to a particular account. It does not follow your child to a different one, and on the teen platforms the linking has to be agreed to rather than imposed. So the setup is a boundary your family has settled on together, which is a real thing and does real work, and it is not a fence.
Take that as an argument for the conversation rather than against the settings. A teenager who agreed to the arrangement and understands why it exists behaves differently from one who had it done to them, and the second one has more reasons to route around it. Set the controls, and put at least as much effort into the reason your child accepts them.
Monitoring Apps Do Not Fill the Gap
Parents often reach for a monitoring tool at this point, and it is worth knowing what those do before you rely on one. They were built to watch messaging and browsing, and they do not read what is inside an AI conversation. Installing one and assuming the AI side is now covered leaves you with a false sense of security, which is a worse position than knowing the gap exists.
Device-level tools still earn their place for the things they were made for, like keeping a device out of a bedroom overnight or limiting when it works. Use them for that. Just do not expect them to tell you what your child asked a chatbot on Tuesday afternoon.
Not All Chatbots Are the Same Thing
One distinction does more work than any setting. A homework assistant your child asks about long division is built for a task and ends when the task does. A companion chatbot is built for open-ended personal conversation, designed to keep your child talking and to encourage attachment, and that is a different proposition whatever the age rating says.
So look at what an app is for before you decide how you feel about it. The design intent tells you more than the marketing does, and it is the question worth asking about any new app that arrives on your child's device this year.
What Works Better Than Settings
Put the device somewhere shared. That single decision does more than every switch in this article combined, and it costs nothing. Beyond that, the durable protection is the one your child carries themselves, which is knowing that these systems produce confident wrong answers and being in the habit of checking. That is covered in teaching kids to use AI well, and it outlasts any control panel.
Keep the door open too. A child who thinks they will be in trouble for mentioning something odd will not mention it, and you have then lost the only monitoring that ever really worked. If you are still deciding the ground rules rather than the settings, whether your homeschooler should use AI for schoolwork is the place to start. And if your worry is about how your child is doing rather than about which switch to flip, that is a conversation for your doctor, not for a settings page.
Once the tech questions are settled, the rest of the year still needs a plan.
Get the GuideA step-by-step plan for your first weeks at home.
A Note from Homeschool Teacher Guide: What This Really Means for You
We will not tell you a setup keeps your child safe, because no honest article can. What the controls do is narrow the ways things go wrong, and that is worth ten minutes of your evening. What they do not do is tell you what your child is thinking about, which is the thing you wanted to know when you started reading.
The part we would hold onto is the shared room and the open door. Homeschooling gives you more of both than most parents get, and neither one expires when a company changes its settings page. Set the controls, check them again in a few months because they will have moved, and spend the rest of your effort on the conversation.
Did this save you hours of research?
We keep these guides free and updated. If this one helped, you can say thanks.
☕ Buy us a coffee ($5)Frequently Asked Questions
What Controls Exist?
It differs by company and changes often. Checked August 2026: OpenAI lets you link your account to a teen's and then set quiet hours and switch off voice mode, memory, image generation and model training. Google lets a parent enable Gemini for an under-13 through Family Link, with image generation blocked under 18. Anthropic requires users to be 18, so nothing needs configuring.
Will They Show Me What My Child Talked About?
Almost never, and this surprises most parents. The platforms hand you switches rather than transcripts, with no conversation access and no topic summaries. You can shape when and how your child uses the tool while learning very little about what was said, which makes the settings a boundary rather than a window.
Can My Child Get Around Them?
The controls attach to an account rather than to your child, and linking has to be agreed to rather than imposed. That is not a reason to skip the setup, since a boundary your child accepted still does real work. It does mean you should not treat the settings as a fence, or stop paying attention once they are on.
Does a Monitoring App Cover This?
Not usually. Those tools were built for messaging and browsing and do not read what is inside an AI chat. Installing one and assuming the AI side is handled gives you a false sense of security. Use device-level tools for what they are good at, like keeping a device out of a bedroom overnight.
Are All Chatbots the Same Risk?
No. A homework assistant is built for a task and finishes when the task does. A companion chatbot is built for open-ended personal conversation and designed to keep your child talking, which is a different proposition whatever the age rating says. Check what an app was made for before deciding how you feel about it.
Sources
This guide was written in August 2026 from the following sources. Parental controls on these platforms change several times a year, so confirm current settings with the provider before relying on anything here.