- ChatGPT for Teens is designed for users ages 13 to 17, but OpenAI says users do not need to verify their age to remain in the protected experience. [1] - The product adds Study Mode, quizzes, Learning Visualizations and “responsible homework reminders” that redirect apparent attempts to shortcut assignments. [2] - Parents can set Quiet Hours, disable memory or voice mode, remove image generation, opt out of model training and receive limited alerts about high-risk situations. [3] - OpenAI’s age-prediction system uses account and usage signals, while adults incorrectly placed in the teen experience can verify their age through Persona.
[4] OpenAI on Tuesday, August 18, launched ChatGPT for Teens, a version of its consumer chatbot that automatically applies additional safeguards when users say they are between 13 and 17 or when the company’s systems estimate that an account belongs to someone under 18. [1]
The rollout combines stricter handling of self-harm, eating disorders, dangerous activities, violence and sexual content with features intended to make ChatGPT function more like a study aid than an answer engine. OpenAI also says the teen experience will avoid romantic language, discourage emotional dependence and avoid implying that it has feelings or consciousness. [2]
A protected experience without mandatory age verification #
OpenAI says ChatGPT for Teens is applied automatically to eligible accounts. Its age-prediction system can examine general topics discussed, usage times, how and when an account is used, and the account’s age. The company acknowledges that the system can make mistakes and gives adults a way to remove the protections through age verification. [4]
The current process is therefore an age-assurance system rather than a universal identity check. Users do not need to verify their age to remain in the teen experience. Adults who want the protections removed can be directed to Persona, which may request a live selfie, government identification or both, depending on the country. OpenAI says it receives age-related confirmation rather than the uploaded identification, while Persona deletes the selfie or ID within seven days after verification. [4]
OpenAI says parents do not receive access to linked teens’ conversations. Instead, a parent or guardian can manage selected settings after the teen accepts an account-linking invitation. Available controls include Quiet Hours, disabling voice mode, turning off memory, removing image generation and opting out of the use of the teen’s conversations to improve OpenAI’s models. [3]
Homework help built around friction #
The teen product includes Study Mode, which uses guiding questions and step-by-step explanations, along with quizzes and Learning Visualizations. OpenAI has also added Study Hours that keep Study Mode on by default during selected periods. [2]
A new Responsible Homework Reminder is intended to recognize when a teen appears to be seeking a quick answer or trying to bypass an assignment. In those cases, ChatGPT for Teens can redirect the interaction toward collaborative problem-solving rather than immediately producing a solution. OpenAI says the feature is designed to support learning outside school, while its separate ChatGPT for Teachers product remains aimed at institution-managed classroom use. [2]
Racquel Gibson, a high school math teacher in Miami, quoted in OpenAI’s announcement, said the homework feature could help students understand mistakes and continue working when no adult is available to help. [2] OpenAI also announced a partnership with CodeAI focused on helping students and educators understand how AI works, question its outputs and use it in classrooms with teachers involved.
[2]Robbie Torney, senior director of AI programs at Common Sense Media, previously described OpenAI’s parental controls as “a good starting point” and said they work best with family rules, conversations about responsible use and active parental involvement. [3] Both comments come through OpenAI’s materials or reporting about the company’s product, rather than an independent evaluation of the teen experience.
Safety alerts remain deliberately narrow #
Parents with linked accounts can receive notifications in limited high-risk situations, including indications that a teen may need urgent offline support. OpenAI says its system can flag potential self-harm; a specially trained team reviews the situation, and parents may be contacted by email, text message and push notification unless they opt out. The company says it is expanding notifications to include some eating-disorder-related situations and alerts when a linked account is banned for violent activity, while excluding ordinary fictional writing, gaming, news discussion, political discussion and general anger from those notices. [3]
OpenAI says the safeguards are designed to preserve teen privacy by sharing only information considered necessary for safety intervention. That approach leaves parents without routine access to a teen’s chats, and it means the alert system is not intended to provide comprehensive monitoring. [3]
That limitation matters because a January 30, 2026 academic preprint evaluating OpenAI’s parental-control system found that notifications were selective rather than comprehensive in the tested setup. The researchers reported no alerts for several tested categories, including fraud, hate speech and malware, while physical-harm, pornography and some health prompts produced intermittent alerts. The study examined a specific system configuration and does not establish how the August 18 launch performs, but it illustrates why parental controls should not be treated as a complete safety net. [5]
OpenAI says it is publishing new under-18 evaluations in its system cards covering self-harm, eating disorders, violence, age-restricted goods and services, and sexual content. The company has not disclosed a launch-wide accuracy rate for the age-prediction system or the new teen safeguards. [2]
Companies mentioned #
Further sources #
[[1] OpenAI announced ChatGPT for Teens on August 18, 2026, for users who identify a… ↗](https://openai.com/index/chatgpt-for-teens/)
[[2] OpenAI described Study Mode, responsible homework reminders, quizzes, Learning … ↗](https://openai.com/index/chatgpt-for-teens/)
[[3] OpenAI’s parental-controls documentation describes account linking, conversatio… ↗](https://openai.com/index/introducing-parental-controls/)
[[4] OpenAI’s age-prediction documentation explains the signals used, automatic teen… ↗](https://help.openai.com/en/articles/12652064-age-prediction)
[[5] The January 30, 2026 arXiv preprint evaluated OpenAI’s parental-control system … ↗](https://arxiv.org/abs/2601.23062)
The stories that matter, in one email. Free — unsubscribe anytime.