OpenAI has announced the launch of a new version of its ChatGPT chatbot aimed exclusively at the 13-17 age group, in a move targeting the first generation that grew up with artificial intelligence.
This initiative comes at a time when teenagers are increasingly turning to artificial intelligence technologies to perform their schoolwork, obtain daily advice, and even treat these systems as a digital companion.
The new version features strict security controls that prevent the robot from displaying content related to self-harm, suicide, or romantic and sexual relationships.
The robot's academic features are also designed to help students understand their lessons, rather than providing ready-made answers or essays that can be used for cheating.
The company says the goal is to establish healthy AI usage habits that are appropriate for the age group of teenagers.
In this context, Ann O’Leary, Vice President of Global Policy at OpenAI, explained that the company seeks to deal with teenagers in a manner appropriate to their mental maturity, without belittling them or exposing them to inappropriate content.
Education experts, parents, and teachers believe that teenagers' use of smart chatbots carries significant risks, as these technologies have previously been accused of causing cases of academic cheating, and even of affecting the mental health of teenagers, leading to cases of suicide.
Although adults can also fall into the trap of emotional engagement with artificial intelligence, teenagers are more vulnerable to this risk due to the incomplete development of their brains.
Previous research has revealed that ChatGPT provided detailed advice to teenagers on how to use alcohol and drugs, conceal eating disorders, and even wrote poignant suicide notes on demand.
According to an American study published in 2025, more than 70% of teenagers use chatbots for companionship, and half of them do so regularly.
OpenAI stated in its announcement on Tuesday, August 18, that the teen robot is programmed not to suggest that it has feelings towards the user, or that it is aware or experiencing emotions, in addition to completely banning emotional or sexual conversations.
The company is also working on studying the phenomenon of "excessive emotional dependence" on these technologies, which its officials consider common among young people.
OpenAI does not verify users' identities, but uses age estimation techniques based on the nature of the questions asked. If a user is determined to be under 18, they are automatically moved to the version designed for teenagers, a mechanism similar to that used by Meta for Instagram accounts targeting teenagers.
As for parental controls, they are optional and allow parents to set "quiet" times during which access to the application is prohibited, and to receive notifications in dangerous situations, such as the possibility of the teenager harming himself.
The new notifications include warnings related to eating disorders, with a focus on directing the user to external support when needed.
On the educational side, the new robot is designed to guide students towards understanding the course material on their own, rather than to provide direct answers.
OpenAI is developing educational versions for teachers and plans to expand interactive learning, based on studies that confirm students learn better when they face challenges on their own.
