OpenAI has begun rolling out ChatGPT for Teens, a stricter version of its chatbot that will be automatically applied to users the system identifies as being under 18, without requiring them to create a new account. The move follows the company's announcement last September, made after it was sued over allegations that ChatGPT contributed to the death of 16-year-old Adam Raine, that it would build a system to detect and restrict teenage users.

The update is built around learning and safety. On learning, it bundles existing tools such as Study Mode, which guides users to answers via questions rather than giving them outright, alongside new features including "Study Hours" that automatically activate Study Mode at set times chosen by teens or parents, and expanded visual explanations now covering more than 300 topics, up from 70 at launch. On safety, OpenAI is restricting the chatbot from describing itself as a friend or implying it has feelings, increasing the frequency of break reminders, and adding parental notification for policy violations related to eating disorders, building on notifications already sent for violent-threat violations.