OpenAI on Tuesday launched ChatGPT for Teens, a dedicated experience for users aged 13 to 17 that automatically routes anyone identifying as under 18, or flagged by the company’s age-prediction system, into a restricted mode. Under a new under-18 model spec, the product blocks romantic and sexual chats, layers interventions across self-harm, violence and eating disorders, and bars the model from using romantic language, encouraging emotional dependence, or implying that it has feelings or consciousness.

The framing is behavioral, not just topical. “We went through and identified what are the hypothetical cues that a model could give that might make a teenager kind of develop a relationship to it,” said Allison Mishkin, OpenAI’s head of child development. That’s the tell: the concern isn’t only what the model says, but what kind of attachment it invites.

A Study Mode walks users through problems with guiding questions instead of direct answers, homework reminders surface when the system suspects cheating, and break prompts appear after 90 minutes of activity within a three-hour window. Parents who link accounts can set Quiet Hours, manage settings, and receive safety notifications in limited high-risk situations.

The rollout arrives under legal pressure. Florida sued OpenAI and chief executive Sam Altman in June, accusing the company of failing to adequately address safety risks for minors. In April, seven families in Tumbler Ridge, British Columbia, sued OpenAI and Altman over conversations preceding a school shooting. OpenAI has disputed the claims. Further age-assurance guidance is expected later this year.

ChatGPT reached 900 million weekly users before any teen-specific safeguards existed. Tuesday’s product is what youth-safety architecture looks like when it’s retrofitted onto a platform after the teenagers have already arrived.

Sources