OpenAI is launching a new version specifically designed for users under 18, featuring stricter security measures and parental controls. Called “ChatGPT for Teens,” this system allows parents to link their children’s accounts to their own and manage their usage habits.

The new security system automatically activates if a user is detected as being under 18 or indicates they are between 13 and 17 years old. With this step, OpenAI aims to help young people develop healthier habits in the digital world and protect their critical thinking skills.

Automatic Security and Restrictions

The system is structured to offer a less human-like experience for young users. In particular, the voice reply feature can be disabled on young accounts, preventing the AI ​​from giving the feeling of talking to a real person.

OpenAI will send notifications to young users reminding them to take regular breaks. These reminders help balance digital usage time by emphasizing that the user is interacting with an AI. Children under 13 are not allowed to use the platform.

Parental controls and education-focused tools

Parents can limit the hours their children can access ChatGPT through the “Quiet Hours” feature. This control system ensures that AI use remains consistent with sleep and daily activities.

Sending notifications to parents in high-risk situations is also becoming a valuable part of the system. Risky requests regarding sensitive issues such as eating disorders are reported to parents within approximately one hour after undergoing human review.

Stating that young people largely use ChatGPT for educational purposes, OpenAI is also developing learning tools. The Study Mode feature focuses on understanding the subject matter step-by-step rather than solving assignments directly, and can now be enabled by default during specific study hours. The system also supports student practice with responsibility-focused assignment reminders, tests, and visualization tools.

Do you think these kinds of restrictions and parental controls for young people are enough to make the use of AI safer?