OpenAI has launched ChatGPT for Teens, introducing a version of its AI chatbot with stronger safety protections for users aged 13 to 17.
The move reflects a growing challenge for AI companies: as younger people increasingly use generative AI for education, creativity and everyday tasks, platforms must find ways to make these tools useful without exposing minors to inappropriate content or encouraging unhealthy relationships with AI.
Under the new system, ChatGPT accounts are expected to default to the teen experience when OpenAI determines that a user is under 18 or when someone identifies themselves as being between 13 and 17 during sign-up. Children under 13 remain prohibited from using the platform.
OpenAI tightens AI safeguards for younger users
The new teen experience goes beyond simply restricting adult content.
OpenAI says its under-18 safety rules prohibit the chatbot from engaging in romantic or sexualised roleplay, using romantic language with teens, encouraging emotional dependence or presenting itself as having human feelings.
The company is also introducing additional controls around how the chatbot interacts with younger users, amid wider concerns about people developing strong emotional attachments to AI systems.
While OpenAI has not detailed every aspect of the changes, the rollout includes customisation features such as accent colours and voice options designed to give teenagers a personalised experience without making the chatbot appear like a human relationship.
The company said the changes are part of its broader approach to building safer AI products for younger users rather than a response to one specific incident.
ChatGPT becomes more focused on learning
Education is another major part of the teen experience.
OpenAI is expanding Study Mode, which is designed to guide students through problems using questions and step-by-step explanations rather than simply providing answers.
The company is also introducing features such as quizzes, learning visualisations and scheduled Study Hours, allowing users to set periods when Study Mode remains active by default.
There are also quiet times, which can be used to temporarily switch ChatGPT off.
These features position ChatGPT less as a shortcut for completing schoolwork and more as an AI learning assistant.
For OpenAI, that distinction could become increasingly important as schools and parents debate how much students should rely on AI for assignments and academic work.
Parents get more visibility
OpenAI is also giving parents additional safety controls through linked accounts.
Certain high-risk conversations, including prompts related to eating disorders, can trigger alerts to parents. OpenAI says these alerts are reviewed by a person before being sent, with the company aiming to notify parents within an hour.
The move gives parents a greater role in monitoring how teenagers interact with AI while attempting to preserve some independence for young users.
OpenAI is also partnering with CodeAI, an education-focused AI company, to help teenagers and teachers understand how AI systems work, question their outputs and use the technology responsibly.
AI companies face growing pressure over younger users
OpenAI is not the only company reconsidering how AI products should work for minors.
Anthropic’s Claude remains restricted to adults, while Google’s Gemini has also faced restrictions around younger users.
The issue is becoming increasingly important as generative AI moves from being an experimental technology into everyday tools used for learning, communication and productivity.
For AI companies, younger users represent a significant future market. But serving that market comes with additional responsibilities around safety, privacy, education and responsible technology use.
OpenAI’s ChatGPT for Teens rollout therefore represents more than a new version of the chatbot.
It signals a broader shift in the AI industry: the companies building consumer AI products are increasingly being forced to design for different levels of risk — especially when their users are children and teenagers.
As AI becomes more embedded in education and everyday life, the ability to balance access, usefulness and safety could become just as important as the technology itself.
