OpenAI launches ChatGPT for Teens with study tools and parental controls

Automated detection and safety controls

OpenAI introduced ChatGPT for Teens on Tuesday, August 18, 2026, targeting users aged 13 to 17 with safety guardrails and parental oversight. The dedicated interface activates automatically when a user registers as a minor or when OpenAI's systems detect an under-18 user through more than 2,000 behavioral data signals, including account age and typical login times. Under the terms of service, children under 13 remain barred from using the platform. OpenAI removed the chatbot's human voice response option for teen accounts to limit personification. The software also introduces prompts urging teenagers to take regular breaks, reminders that ChatGPT is an artificial intelligence, and notices advising caution when uploading private photos.

OpenAI executive Lauren Jonas explained the automatic activation mechanism for minor accounts.

There is no need for a teenager to create a new account or change anything. If we can predict that you are under 18 or if you tell us, this becomes your default experience.

Educational tools and study mode

The company designed the teen interface to support schoolwork through guided learning rather than supplying direct answers. A study mode, developed in collaboration with Stanford University and the Massachusetts Institute of Technology, guides students through broader concepts using follow-up questions and revision notes. ChatGPT for Teens includes automated homework detection that redirects users to study mode if they attempt to copy solutions or bypass exercises. Teens and parents can configure dedicated study hours to turn on study mode by default, alongside quizzes and interactive practice tools. Company data indicates that nine out of ten teenage users access ChatGPT to acquire skills, gather information, or learn.

Parental oversight and risk notifications

Parents who link their accounts can enforce quiet hours, which lock the application during scheduled daily intervals. The system monitors conversations for high-risk topics, including self-harm, eating disorders, violence, and explicit material. Safety alerts are reviewed by human staff before being dispatched to parents by text message, email, or in-app notification, with OpenAI aiming to deliver warnings within one hour. OpenAI has also restricted the chatbot's conversational persona to prevent emotional dependence.

Jonas described the new boundaries preventing the system from mimicking human attachment.

We added tighter restrictions to relationship boundaries. The model should not call itself a friend, suggest it has personal feelings, or suggest sentience.

Legal scrutiny and classroom partnerships

The global rollout, scheduled to conclude across all supported markets over the next two weeks, arrives while OpenAI prepares for an initial public offering at an $852 billion valuation. The company faces multiple lawsuits alleging its platform harmed children, including a 2025 filing by the parents of 16-year-old Adam Raine following his death by suicide. In June 2026, Florida became the first US state to sue OpenAI, alleging the startup misrepresents platform safety for minors and fosters addiction. Academic research remains divided on classroom impact; a 2025 study in the Proceedings of the National Academy of Sciences found that while AI helped students solve math exercises, test scores declined in subsequent exams without the tool. OpenAI continues to establish public education partnerships in Estonia, Greece, and Slovakia, while offering free access to teachers in the United States.