OpenAI unveils ChatGPT for Teens with stronger guardrails to tackle safety risks

OpenAI has introduced a new version of its AI chatbot, ChatGPT for Teens, equipped with stronger safety measures and parental controls, amid growing concerns over the potential risks associated with AI chatbots. The new feature is designed to limit conversations on high-risk topics such as self-harm, violence, and eating disorders, while providing educational tools to assist students in learning rather than simply providing answers.

When the system detects a user under 18 or self-identifies as being between 13 and 17 years old, it automatically engages and allows parents to set quiet hours, manage settings, and receive notifications when sensitive topics are discussed. This initiative follows in the footsteps of other tech giants like Meta, who have also enhanced protections for teenagers on their platforms due to mounting legal challenges and evidence of mental health issues among young users.

With the increasing global push to restrict social media use for teens due to potential harm, OpenAI faces additional pressure. For OpenAI, which has already been sued for alleged harm to children caused by ChatGPT, the challenge is further complicated by the growing trend of treating chatbots as confidants offering emotional support, for which AI lacks the necessary human understanding.

OpenAI emphasizes its belief in using AI to expand educational opportunities, hence the development of the ChatGPT for Teens mode, which includes features like a study mode, homework reminders, quizzes, and visualizations. According to a recent Pew Research Center report, U.S. teenagers primarily use AI for searching information, getting homework help, and entertainment.

OpenAI utilizes a combination of self-reported age, behavioral signals, account details, and verification to estimate a user's age and defaults to the safer setting when unsure.