OpenAI’s junior version of ChatGPT with guardrails has launched

OpenAI has introduced a more restrictive version of its ChatGPT chatbot, designed specifically for teenagers aged 13-17, in response to concerns surrounding the potential misuse of AI by younger users. The new "ChatGPT for Teens" model incorporates several safeguards to prevent conversations that could be harmful, such as those related to violence, self-harm, sex, and mental disorders. Parents are alerted if their child engages in restricted topics when parental controls are enabled.

Upon signing up, users must state their age, although OpenAI does not employ an age-verification tool. Instead, the chatbot analyzes user inputs to determine if a user is a young teenager; if so, the model automatically activates teen mode. In terms of features, the teen version includes a "Study Mode," which allows young users to take quizzes and visualize subject-related content. Parents can monitor their child's study habits using the "Study Hours" feature, which keeps track of the time spent on homework.

OpenAI emphasizes the importance of providing AI access to teenagers while ensuring their safety and promoting healthy online habits. Studies indicate that up to 70% of young users engage with chatbots for companionship, which can be problematic given the vulnerability of teenagers' brains to potential risks associated with synthetic media and bad advice. OpenAI's CEO, Sam Altman, has acknowledged the company's ongoing efforts to study "emotional overreliance," particularly concerning younger users.

To maintain a healthy balance, the platform encourages real-world relationships and healthy habits through built-in break reminders and clear identification of the AI component. OpenAI acknowledges the potential for forming unhealthy attachments to AI and has designed the "ChatGPT for Teens" to acknowledge its artificial nature.