ChatGPT parental controls: the full 2026 setup guide for families
You came for the settings. Once they're on, the part the controls can't do is knowing whether tonight's homework was actually understood — paste the finished work and find out in twenty seconds.
ChatGPT's stated minimum age is 13, with parental permission required under 18. Parental controls work by linking a parent account to a teen account, then switching off memory, chat-history training, image generation and voice as needed. There is no separate kids' plan — the controls plus a shared house rule are the setup.
Most parents arrive at this page after finding ChatGPT already open on a laptop. The good news is that the controls exist and take about ten minutes. The important part is knowing which ones matter and which ones give a false sense of safety.
The ten-minute version
- Give the teen their own account with their real birth year — that is what turns on the teen safeguards.
- Link it to your account from Settings → parental controls, and accept the invite on their device.
- Turn off model training on their chats, then turn off memory.
- Decide voice mode and image generation deliberately rather than leaving the defaults.
- Set quiet hours so late-night homework sessions are not an option.
- Write one house rule about schoolwork — the controls do not cover that part.
The age rules, stated plainly
The setting that doesn't exist
Check tonight's homework in twenty seconds.
No toggle tells you whether they understood it. Paste the finished work or snap a photo — first check is free.
- 13 is the stated minimum age to hold an account.
- Under 18 requires parental permission according to the terms of use.
- Under 13 is not permitted — accounts are meant to be for a parent, used together.
- Nothing verifies age at signup, which is why the household rule matters more than the policy.
The longer version of the age question, including what to do for a 9-year-old who wants access, is in our guide to how old you have to be to use ChatGPT.
Setting up parental controls, step by step
- Create the teen's account with their own email, using their real birth year — this is what activates the teen safeguards.
- From your own account, open Settings, then the parental controls or family section, and send an invite to the teen account.
- Accept the invite from the teen's device. The link is what lets your settings apply to their account.
- Turn off model training on their conversations. This is the single setting most parents want changed.
- Turn off memory so the assistant does not carry personal details between chats.
- Decide on voice mode and image generation. Both are fine for most teens; both are the features younger kids use unsupervised for the longest.
- Set quiet hours if the option is available on your plan, so late-night use is not an option.
Device by device: where each setting actually lives
The ChatGPT controls sit in the account, so they follow your child to every device. The limits that stop a phone being used at 1 a.m. sit in the device instead. Most families need both halves.
iPhone and iPad
- In the ChatGPT app: profile icon → Settings → Parental controls → accept the invite sent from your account.
- Settings → Data controls → turn off "Improve the model for everyone".
- Settings → Personalisation → turn off Memory.
- iOS Settings → Screen Time → App Limits: add ChatGPT with a nightly cutoff.
- iOS Settings → Screen Time → Content & Privacy Restrictions: block new app installs so a second assistant does not appear next week.
Android phone or tablet
- Same in-app steps: profile icon → Settings → Parental controls, then Data controls and Personalisation.
- Google Family Link → your child → App limits: set a daily cap and a bedtime for ChatGPT.
- Family Link → Controls → Google Assistant/Gemini: decide this separately, because Gemini is on the device whether or not ChatGPT is. See is Gemini safe for kids.
Laptop or desktop browser
- Sign in as the teen, then check Settings → Data controls and Personalisation match what you set on the phone. They should, but confirm once.
- Leave chat history on. It is the record that helps if a teacher ever queries a piece of work.
- Keep the laptop out of the bedroom for homework. It does more than any toggle.
School Chromebook
- The district controls this device, not you. Your settings do not apply to a school-managed account.
- Ask the school which assistants are enabled and under what policy — your school's AI policy, explained has the four questions to email.
Windows PC
- Copilot is built into Windows and Microsoft 365, so ChatGPT settings are only part of the picture: is Microsoft Copilot safe for kids.
- Microsoft Family Safety handles screen time and app limits at the account level.
Working through the whole house? Our parental controls setup guide covers every device and app in one pass, and the printable ChatGPT setup checklist is the tick-box version.
Is there a ChatGPT family plan?
There is no dedicated family or kids' subscription. There are business and team plans built for workplaces, which are not designed for households and are more expensive per seat than they are worth for most families. In practice the family setup is: separate accounts, linked parental controls, and one written house rule about schoolwork.
What the controls do not cover
Controls limit features. They do not tell you whether the homework that came out of the chat was understood. That is the gap most parents actually feel — the assignment is finished, correct, and the child cannot explain a line of it.
That is a conversation, not a setting. Our family AI agreement template covers the schoolwork boundary in one page, and the AI homework rulebook breaks the line down by grade and assignment type.
The house rules worth writing down
- AI can explain, quiz, and check. It does not write, solve, or finish.
- Anything AI helped with gets read out loud and explained in their own words before it is handed in.
- No personal details in prompts: no full name, school, address, or photos of other people.
- Anything strange, upsetting, or too grown-up gets shown to a parent, with no consequence for showing it.
Other assistants your child probably has access to
ChatGPT is rarely the only one. Gemini is built into Android and Google Workspace for Education, Copilot ships inside Windows and Microsoft 365, and Snapchat's assistant is in an app most teens already use. Each has its own controls: see is Gemini safe for kids, is Microsoft Copilot safe for kids, is Snapchat My AI safe for kids, and is Character.AI safe for kids.
Doing this in the first weeks of school
Term time is when the settings actually get tested. Two things are worth doing in the first fortnight: leave chat history on, so there is a record of how a piece of work happened if a teacher ever queries it, and read your school's AI policy before you write your own rule — the two need to agree. Your school's AI policy, explained walks through the four questions to email a teacher, and set your family's AI homework rules before school starts is the conversation version.
The check the controls cannot do
No toggle tells you whether your child understood the assignment they just handed in. Paste the finished work into our free AI homework checker and you get a plain-language read on what they actually grasped — no accusation, no detector score. If the problem is the other direction and you need tonight's topic explained without handing over the answer, use explain my kid's homework.
Frequently asked questions
Does ChatGPT have parental controls?
Yes. A parent account can be linked to a teen's account, which lets you manage model training on their chats, memory, voice mode, image generation and quiet hours. The controls apply to the linked teen account, not to a shared adult account.
How old do you have to be to use ChatGPT?
The stated minimum is 13, and users under 18 need parental permission. Under-13s are not permitted their own account, so younger children should use it alongside a parent on the parent's account.
Is there a ChatGPT family plan?
No. OpenAI offers individual, team and business plans, but no dedicated family or child subscription. Families typically use separate accounts with linked parental controls plus a written household rule about schoolwork.
Can I see my child's ChatGPT conversations?
Linked parental controls manage features and safety settings rather than giving you a transcript feed. If you want visibility into schoolwork specifically, the practical approach is asking your child to explain the finished work out loud, or running it through a parent-side homework check.
What is the first ChatGPT setting I should change?
Turn off model training on their conversations, then turn off memory. Those two cover the data concerns most parents have. Voice and image generation are worth deciding on separately based on your child's age.
How do I set up ChatGPT parental controls on an iPhone or Android?
The controls live in the ChatGPT account rather than the device, so the steps are the same on both: link the teen account from your own Settings, accept the invite in the app on their phone, then set training, memory, voice and quiet hours. Device-level screen time limits are separate and are set in iOS Screen Time or Google Family Link.
Do ChatGPT parental controls stop cheating on homework?
No. They limit features and data, not what a child asks the assistant to write. The homework side needs a household rule plus a way to confirm understanding — asking the child to explain the finished work out loud, or running it through a parent-side homework check.
Have a specific piece of homework in mind?
Paste it — or a screenshot of it — into the Homework Check on our homepage and ask your question. You'll get a plain-language read on whether the thinking looks like your kid's, what stood out, and the questions worth asking them. First check is free.