ChatGPT for Teens is rolling out as the default experience for accounts OpenAI predicts belong to someone under 18. OpenAI’s parent-resources page says teens do not create a new account. If the system estimates they are under 18, or they already stated they are 13 to 17, the teen tools and safeguards turn on automatically.
How ChatGPT for Teens gets assigned
OpenAI’s age-prediction help article says the model looks at account-level signals. Those include how long the account has existed, typical times of day, usage patterns, and stated age. The check can still apply even if someone already entered a birthday. Therefore an adult can land in the teen experience by mistake. Adults can verify age in Settings, including a selfie check through Persona. The same help page says the EU rollout follows in the coming weeks. In Italy, a prompted user has 60 days to finish verification.
OpenAI says the account stays active. Teens can still learn, create, and ask questions. The company also says it will not show ads in those accounts. No system is perfect, and the help article says so in those words.
Study tools, then the safety defaults
Study Mode still walks a student through a problem instead of dumping the answer. Study Hours can turn Study Mode on during set times. Teens can set those hours, or a linked parent can. Responsible homework reminders try to catch a shortcut and steer the chat back into Study Mode. Official family pages also list quizzes, flashcards, projects, and interactive math and science visuals. OpenAI says 18 million weekly users already touch those interactive lessons.

Safety defaults sit on top of that. Break reminders fire more often and stress that ChatGPT is an AI, not a friend waiting for a reply. Quiet hours can shut the chatbot off. Parents with a linked account can still tighten features, including Study Mode on every new chat. OpenAI already told linked parents when a teen is banned for violent threats. We covered that parent-alert expansion last month. The new teen package keeps those alerts in the family hub.
What this does not settle
OpenAI has not published a hit rate for age prediction. Clever teens will still try to look older. Therefore treat this as a default experience, not a lock. For another 2026 family-control beat, see Threads parental supervision tools. If you run a family account, start on the official parent-resources page and confirm the teen’s settings after the next app update.