OpenAI has activated an age prediction system on ChatGPT to assess whether users are youth under 18 years old, as a measure to ensure youth safety.
OpenAI's official website announced the activation of an age prediction system on ChatGPT for general users to evaluate whether an account belongs to a youth under 18. The goal is to enhance safety and tailor the user experience appropriately by age, following guidelines outlined in the adolescent safety roadmap.
This ChatGPT system operates through an age prediction model that analyzes behavioral data and account-level information, combining factors such as account age, daily activity patterns, and the age provided during registration.
When the system estimates that an account likely belongs to a user under 18, ChatGPT automatically activates additional protective measures to limit access to sensitive or risky content, such as graphic violence, harmful activities or behaviors, sexual role-playing or violence, self-harm content, and materials promoting unhealthy beauty standards or dieting. These measures are based on child development research and psychological expert advice to address adolescents’ emotional regulation and risk perception differences.
For users over 18 who are mistakenly classified as youth, they can verify their identity to restore full access rights by submitting a selfie to confirm their age.
Users can check their account safety status and initiate the verification process at any time via the account settings menu.
Additionally, OpenAI has introduced parental controls that allow guardians to customize their children’s usage, such as setting usage time limits, managing data retention and training usage, and receiving alerts if signs of severe distress are detected during use. This age prediction system is being gradually rolled out worldwide and will begin expanding to the European Union region within the next few weeks.
. Source:OpenAI