OpenAI has launched a version of ChatGPT designed specifically for teenagers, adding stricter safety defaults, study-oriented features and controls for families. The release arrives years after young people began using the service and amid lawsuits and public concern about chatbot interactions involving mental health and suicide.
ChatGPT launched in late 2022 and has since reached 900 million weekly users, according to the company. OpenAI’s new approach acknowledges that a general-purpose product with an enormous audience needs a distinct experience for minors rather than relying only on broad rules written for everyone.
Study Mode tries to shift the product from answers to learning
The teen experience includes Study Mode, which is designed to guide students through questions and steps instead of simply producing a finished answer. It can provide reminders about homework cheating and use quizzes and visualisations to support learning. Parents can make Study Mode the default setting.
That design addresses a tension that has followed generative AI into classrooms. A tool can explain a difficult concept, but it can also complete an assignment without helping the student understand it. Prompting a young person to work through a problem may improve the educational value, although the source does not offer evidence that the feature produces better learning outcomes.
It is also unclear how difficult the restrictions will be to evade. A default can shape normal use, but determined users may try alternative instructions, accounts or services. The difference between a helpful educational nudge and a dependable safeguard will depend on how the system behaves outside ideal demonstrations.
Family controls expand, while sensitive conversations remain the test
OpenAI says its protections are based on principles in its Under-18 Model Spec. The company is also adding family controls, safety notifications and Quiet Hours. These mechanisms can give parents more visibility and help establish boundaries around use.
The launch comes against a more serious backdrop than homework assistance. OpenAI faces legal claims concerning chatbot safety and reports that have connected AI conversations with vulnerable teenagers to self-harm or suicide. The existence of a teen product does not resolve those cases, and TechCrunch’s reporting does not establish that the new controls would have prevented any specific harm.
Safety notifications may help surface concerning behaviour, but they also raise practical questions. Families will need to understand what triggers an alert, how reliably the system distinguishes distress from ordinary discussion and what happens after a warning appears. Excessive alerts could be ignored, while missed signals could create misplaced confidence.
A late safeguard can still shape a very large audience
OpenAI has already introduced ChatGPT for Teachers and is working with CodeAI, extending its effort to position the service inside education. ChatGPT for Teens adds a student-facing component to that strategy. Together, these products suggest the company wants to define separate experiences for the people assigning work, completing it and supervising technology use at home.
The timing remains notable. Teenagers have had access to ChatGPT for several years, and the service reached massive scale before these dedicated protections appeared. That delay means the company must now change established habits rather than setting expectations before adoption.
The new experience should therefore be assessed on observable behaviour, not the number of controls in a menu. Study Mode needs to resist becoming an easily bypassed answer generator. Family features need to provide useful signals without pretending software can replace attentive adults or professional support. Safety defaults need to hold up during emotionally complex conversations, where a generic refusal may be as unhelpful as unrestricted advice.
ChatGPT for Teens is a recognition that minors require different product decisions. It is also an admission that the original one-size-fits-all service left an important population without purpose-built protection. The launch creates a framework; independent evidence and real-world use will determine whether it creates meaningful safety.
