OpenAI announced the launch of 'ChatGPT for Teens' on August 19, aimed at reducing the emotional dependency of adolescents on AI while enhancing their learning outcomes. The system automatically applies this version when it estimates a user to be under 18 or when a user identifies themselves as aged 13 to 17.
The product is designed to encourage teens to understand concepts and solve problems independently rather than simply seeking answers through AI. Students preparing for exams can request explanations of unfamiliar concepts or work through practice problems step-by-step to assess their understanding.
If the system detects that a user is attempting to find answers directly, ChatGPT will guide them into 'Study Mode,' encouraging participation in the problem-solving process. This mode was developed in collaboration with teachers, scientists, and education experts, featuring guiding questions, step-by-step learning support, thought process checks, and comprehension verification to assist students in solving problems rather than providing immediate answers.
Additional features include a 'Responsible Task Reminder' that nudges users toward Study Mode if they attempt to skip assignments, quiz and learning visualization tools, and a 'Study Hours' function that activates Study Mode by default during times set by teens or their parents.
Safety measures tailored for teen users are also implemented. Enhanced protective measures at the model and product levels aim to reduce exposure to harmful or developmentally inappropriate content in high-risk areas such as self-harm, violence, eating disorders, risky behaviors, and explicit sexual or violent content. These standards are based on OpenAI's 'Principles for Users Under 18,' which reflect developmental science research and expert opinions.
For accounts linked to parents, they can set usage limits and manage certain settings. Parents will receive notifications if high-risk situations arise.
However, OpenAI emphasized that while AI will assist with learning outside the classroom, the use of AI in educational settings should be led by teachers.
Features to promote healthy and balanced usage are also included. Users will receive reminders to take breaks after extended use, and the system will continuously inform users that they are interacting with AI to prevent them from mistaking it for a human. According to guidelines for users under 18, ChatGPT is designed not to express romantic feelings or encourage emotional dependency, nor to imply that it possesses emotions or consciousness.
Additionally, users are cautioned against sharing personal photos before uploading sensitive images, and initial guidance for teen users introduces learning and safety features. While simple customization options like color and voice are available, the design ensures that useful tools do not blur the boundaries of human relationships.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.