The system does not rely on a single detail. Instead, it looks at everyday signals such as how long an account has existed, when it is usually active, and patterns of user behavior. By reading these clues together, the software makes an educated guess about a user’s age.

If the system wrongly labels an adult as underage, users will have the option to prove their age through an identity verification service. This process will require a live selfie and an official identity card.

OpenAI first announced the development of this system in September 2025. The main goal is to strengthen protection for minors by limiting their exposure to sensitive or potentially harmful content generated by AI chatbots.

However, questions remain about how accurate age prediction can truly be, especially on a platform with nearly 800 million users. Estimating age based on digital behavior is a complex task, and even small errors could affect millions.

AI companies, including OpenAI, have faced heavy criticism in recent years, with some accusing their technologies of failing to adequately protect children. In response, OpenAI announced additional parental controls for ChatGPT in 2025, signaling a broader effort to place safety before speed.