Your teenager’s ChatGPT is now a different product
On August 18, 2026, OpenAI launched ChatGPT for Teens — and it isn’t an app you download or a plan you buy. If someone tells ChatGPT they’re 13 to 17, or OpenAI’s system estimates they’re under 18, they get moved into it automatically. The study tools change, the model’s personality changes, and parents get new controls. Here’s what’s actually different, what parents can set, and where the limits of a settings screen are.
ChatGPT for Teens is a default experience for under-18 accounts, not an opt-in. It leans the product toward learning (Study Mode, homework nudges, scheduled study hours) and away from companionship (no romantic language, no claims of feelings), and it hands parents a small set of real switches.
How does an account end up in it?
Two routes, and only one of them involves anybody choosing:
- Stated age. If the person says they’re between 13 and 17, they’re placed in ChatGPT for Teens.
- Estimated age. If OpenAI’s system estimates the user is under 18, they’re placed in it too — this builds on the age-prediction work OpenAI has been rolling out.
That second route is the one worth understanding. It means an adult account can, in principle, be read as a teen account, and it means a teen who lied about their birthday isn’t automatically out of reach. Neither is a guarantee in either direction — estimation is a signal, not an ID check.
What changes on the learning side
OpenAI’s framing is that ChatGPT for Teens should help outside the classroom without doing the work for the student. Four pieces carry that:
| Feature | What it does |
|---|---|
| Study Mode | Guiding questions, scaffolding and knowledge checks instead of a finished answer. Built with teachers and pedagogy researchers. |
| Responsible homework reminders | New. When it looks like a teen is trying to shortcut an assignment, ChatGPT redirects them toward Study Mode rather than handing over the answer. |
| Quizzes & Learning Visualizations | Practice questions answered inside the conversation, plus visual explanations for concepts that don’t land as text. |
| Study Hours | New. A teen or a parent can schedule windows where Study Mode is on by default — start time, end time, days. |
Study Hours is the one to actually notice. Every other feature here depends on a teenager choosing the harder path in the moment; Study Hours makes the harder path the default between 4pm and 6pm on weekdays, which is a very different bet.
Interactive quizzing is not teen-only. It went out to every consumer ChatGPT plan on August 14, 2026 — see the education plugins lesson for who gets what. ChatGPT for Teens bundles it with the rest; it doesn’t gate it.
What changes on the safety side
Protections are on by default for under-18 accounts, drawn from the Under-18 Principles in OpenAI’s Model Spec. In practice that means age-appropriate handling in higher-risk areas: self-harm, violence, eating disorders, dangerous activities, and explicit sexual or graphic content.
The change that gets least attention and probably matters most is about tone. The updated under-18 model spec goes past blocking romantic roleplay: ChatGPT should not use romantic language, should not encourage emotional dependence, and should not imply it has feelings or consciousness. Add to that break reminders during long sessions and product cues that keep saying “this is AI,” and the design intent is clear — push teens back toward real people.
Two smaller additions round it out: sensitive-image upload reminders that warn before a teen shares a private photo, and teen-specific onboarding that introduces the safety and learning features up front. Customisation is deliberately cosmetic — accent colours and voice variations — so the experience can feel personal without feeling like a friend.
What parents can actually control
Parental controls only apply to a linked teen account — the link is the whole ballgame. Once linked, a parent can:
The parent switchboard, as of August 19, 2026
- Quiet Hours — windows where ChatGPT isn’t available.
- Study Hours — windows where Study Mode is on by default.
- Selected settings — including sensitive content handling and data controls.
- Safety notifications — alerts in a limited set of high-risk situations. OpenAI is adding notifications related to eating disorders, while deliberately limiting how much of the conversation is shared.
Note the shape of that last one. Parents get told there may be a problem; they don’t get a transcript. That’s a design choice about teen privacy, and whether you think it’s the right one probably says more about your family than about OpenAI.
These are product settings, not clinical tools. A notification is a prompt to have a conversation, not a diagnosis, and no toggle substitutes for a doctor, counsellor or school when a young person is genuinely struggling. If something in your household is at that level, treat the app settings as the least important thing on your list.
The part that isn’t a feature
Alongside the launch, OpenAI announced a partnership with CodeAI aimed at teaching students how AI actually works — how to direct it, question it, and catch its mistakes — with teachers shaping how it lands in classrooms. It sits next to ChatGPT for Teachers, which stays the institution-managed route for classroom use.
OpenAI also started publishing under-18 evaluations in its system cards, showing how models perform against teen-specific standards on sensitive content. That’s the piece to watch over the next year: claims about teen safety are cheap, published evals you can argue with are not.
Don’t configure this behind their back; it fails the moment they find out. Sit down together, link the accounts, and set one Study Hours window around whenever homework actually happens. Then ask them to test the homework reminder: have them ask ChatGPT to just write something for a real assignment and watch what it does instead. Whatever it does, that’s the conversation you wanted to have anyway.