OpenAI Reveals Methods to Protect Children on ChatGPT

JAKARTA, KOMPAS.com – The developer of artificial intelligence, OpenAI, has revealed methods to protect child users on its AI-based chatbot platform, ChatGPT. According to Sandy Kunvatanagarn, Head of Public Policy for Southeast Asia at OpenAI, the platform features several mechanisms to handle users from various age groups. These range from minimum age restrictions to the use of age prediction systems to determine the level of service security. “Our terms of service state that our tools are not intended for use by individuals under 13 years old,” Kunvatanagarn stated during a meeting in Jakarta on Wednesday (8/4/2026). OpenAI also employs an age prediction system integrated into the platform. This system utilises the capabilities of large language models (LLM) along with various signals to estimate the user’s age. “If our model considers you to be under 18 years old, we will set everyone by default to a safer model, a safer version of ChatGPT,” Kunvatanagarn explained. Conversely, if the system estimates that the user is over 18 years old, the user will receive the version of ChatGPT that allows access to more features. The safer version also serves as the default setting for users accessing ChatGPT without logging in or without an account. OpenAI has also added several parental supervision features to ChatGPT. A few months ago, the company launched an age prediction system capable of estimating whether a user is an adult or a young user. “This feature also allows parents to link their ChatGPT accounts with their children’s accounts, especially if they are still teenagers,” said Raghav Gupta, Head of Education for Asia Pacific at OpenAI, on the same occasion. Once accounts are linked, parents can adjust various controls through several setting toggles to ensure a safe user experience. For example, parents can determine whether the image generation feature is permitted, limit the duration of a child’s ChatGPT usage, or set specific times when the service cannot be used via the “quiet hours” feature. These features are designed to keep young users’ ChatGPT usage under supervision and appropriate to their age. This approach aims to ensure that AI usage remains suitable for the user’s age. OpenAI describes these steps as part of efforts to encourage safer and more responsible use of AI technology, particularly for users under 18 years old.