OpenAI has launched ChatGPT for Teens, a version of the service that automatically applies stronger protections and learning-focused features to eligible users aged 13 to 17. OpenAI says people who state that age, or who its systems estimate are under 18, are placed into the teen experience by default.
The product combines several ideas OpenAI had already been building — age prediction, parental controls and Study Mode — with new controls aimed specifically at younger users. The result is less a separate chatbot than a different set of defaults around the same service.
Learning is the main product pitch
OpenAI is positioning the teen experience around active learning rather than answer delivery. Study Mode uses guiding questions and step-by-step support, while new responsible-homework reminders are designed to notice when a user appears to be trying to shortcut an assignment and redirect them toward working through it.
Study Hours let a teen or linked parent choose periods when eligible new chats begin in Study Mode automatically. OpenAI is also bringing quizzes and learning visualisations into the experience where available.
That distinction matters because the company is not promising that ChatGPT suddenly becomes a perfect tutor. It is trying to change the behaviour around how a teenager uses the tool: more prompts to reason through a problem, fewer cues that reward simply pasting in homework and collecting an answer.
The Verge notes that several safety components are consolidations of features OpenAI had already introduced. That is fair, but bundling them into one age-specific experience makes the defaults easier to understand. A parent should not need to know the release history of five separate safety features to work out what a teen account is supposed to do.
The safety defaults are stronger
OpenAI says the teen mode adds age-appropriate model protections and product interventions around areas including self-harm, violence, eating disorders, dangerous activities and explicit sexual or graphic content. Linked parents can set Quiet Hours, manage selected settings and receive notifications in limited high-risk situations.
The company is also adding eating-disorder-related notifications, more frequent break reminders during extended use and warnings before teens upload sensitive images. Teen onboarding is supposed to explain those protections from the start rather than leaving them buried in settings.
OpenAI also says the under-18 behaviour rules are designed to reduce emotional dependence. ChatGPT should not use romantic language with teens, encourage an exclusive emotional relationship or present itself as conscious or capable of feelings. That boundary is reinforced with cues reminding users that they are interacting with AI.
There are smaller personalisation options too, including accent colours and voice variations. Those are deliberately cosmetic: OpenAI says the aim is to make the product feel personal without making the AI feel like a human relationship.
The company says ChatGPT for Teens is the next step rather than a finished safety system. It plans further research, evaluations and product changes as it learns how younger users interact with AI.
For families, the most useful change is the default itself. A 13-to-17-year-old should not have to discover a hidden safety mode, and a parent should not have to configure every protection manually. The harder question — how effective those protections are in messy real-world conversations — will need to be judged over time, not from the launch announcement.
Reporting notes