OpenAI launched a new version of ChatGPT designed for teenagers, adding parental controls and stronger safeguards as concerns grow over the risks of AI chatbots for younger users.
The new “ChatGPT for Teens” mode applies additional restrictions to conversations involving high-risk topics, including self-harm, violence and eating disorders, while also providing tools for schoolwork, according to an OpenAI statement published on Tuesday, August 18.
The mode automatically activates when OpenAI estimates that a user is under 18 or when a user identifies themselves as being between 13 and 17. The company said it uses a combination of self-reported age, behavioral and account signals, as well as verification, to estimate a user’s age and will default to the safer setting when it is unsure.
The launch comes as concerns grow over how teenagers use conversational AI, particularly in situations involving emotional support and other sensitive issues. OpenAI is also facing lawsuits alleging that ChatGPT contributed to harm involving children, adding to wider questions about the safeguards that should apply when minors use conversational AI.
New tools for studying and parental controls
The new version includes Study Mode, which uses guiding questions and step-by-step support instead of immediately providing an answer. It also includes quizzes and Learning Visualizations, while homework reminders can redirect users toward Study Mode when they appear to be trying to bypass an assignment.
A new Study Hours setting allows teenagers or their parents to select periods when Study Mode is activated by default, while the service will also display break reminders and warnings about sharing private or sensitive images.
Parents with linked teen accounts can set Quiet Hours, manage selected settings, and receive safety notifications in certain high-risk situations.
The teen version applies additional protections to content involving dangerous activities and explicit sexual or graphic material. OpenAI’s rules for users under 18 also prohibit romantic language, encouragement of emotional dependence and responses suggesting that ChatGPT has feelings or consciousness.
OpenAI said the new protections build on its existing parental controls, age-prediction system and under-18 safety rules, with the company continuing to evaluate how its models perform in situations involving younger users.
