A screen showing OpenAI's ChatGPT for Teens service, which allows parents to configure usage settings. [Provided by OpenAI]
A screen showing OpenAI's ChatGPT for Teens service, which allows parents to configure usage settings. [Provided by OpenAI]

As social media addiction among teenagers has emerged as a global concern, AI services are now taking concrete steps to curb unchecked use by minors.

OpenAI announced Wednesday the launch of "ChatGPT for Teens," a version of its AI assistant designed to help young users learn more effectively while preventing excessive emotional dependence on AI.

The service activates automatically when the system estimates a user is under 18, or when a user identifies their age as between 13 and 17.

OpenAI said the service focuses on helping teenagers understand and work through problems on their own, rather than simply obtaining quick answers from AI.

A student preparing for an exam, for example, can ask for explanations of unfamiliar concepts and work through practice problems step by step, then check whether they have genuinely understood the material.

When a user appears to be seeking a direct answer to an assignment, ChatGPT guides them into Study Mode, prompting them to engage with the problem-solving process instead.

Additional learning tools include a "responsible homework reminder" that steers users who try to skip problem-solving back into the process, along with quiz features and study visualization tools.

The Study Mode was developed in collaboration with teachers, scientists and education specialists. A new "Study Hours" feature lets teenagers or their parents set a schedule during which Study Mode turns on by default.

The ChatGPT logo [Getty Images]
The ChatGPT logo [Getty Images]

Age-appropriate safeguards are also built into the service. OpenAI has strengthened protections to reduce exposure to inappropriate content, including self-harm, violence, eating disorders, dangerous behavior, and explicit sexual or violent material. On accounts linked to a parent, parents can set Quiet Hours to restrict usage and manage certain settings. Alerts go to parents when high-risk situations are detected.

The service also sends break reminders after extended use and continuously identifies itself as an AI to prevent users from mistaking ChatGPT for a human.

"ChatGPT is designed not to express romantic feelings toward teenagers, encourage emotional dependence, or speak as though it has emotions or consciousness," OpenAI said. "We have built it so that the boundary between a useful tool and a human relationship does not become blurred."

Meanwhile, as social media addiction among teenagers has become a serious social issue, countries including those in Europe are pursuing measures to ban minors below a certain age from using social media platforms. Concerns continue to grow that AI could similarly affect teenagers through excessive use and emotional dependence, and further measures are expected to follow.


sjpark@heraldcorp.com