OpenAI has begun rolling out a new system on ChatGPT that estimates a user’s age and applies extra safety limits when an account appears to be used by someone under 18.
The company said the update is focused on protecting teenagers while leaving the experience largely unchanged for adults. According to OpenAI, the goal is to step in only when there are signs that a younger user may be involved, rather than placing broad restrictions on everyone.
The age-estimation system does not require users to submit documents by default. Instead, it looks at how an account behaves over time. This includes how long the account has existed, how it is typically used, when activity usually takes place, and any age details the user has provided.
If the system suggests an account likely belongs to a minor, ChatGPT automatically restricts certain types of content. These limits apply to graphic violence, sexual or violent role play, material related to self-harm, risky online challenges, and content that could encourage unhealthy body image or extreme dieting.
OpenAI said users who state they are under 18 during sign-up already receive these protections. The age-estimation tool is mainly intended for situations where age information is missing or unclear, with the system designed to default to safer settings in those cases.
The company acknowledged that the system will not always be accurate. Users who believe they have been wrongly restricted will be able to confirm their age through a selfie-based verification process handled by a third-party identity verification service. This option can be accessed through the account settings.
OpenAI said the model will continue to be adjusted as it learns which signals improve accuracy. Until then, the system is designed to err on the side of caution.
Alongside the automated limits, OpenAI is also expanding parental control options. Parents can set quiet hours, manage features such as memory or model training, and receive alerts if the system detects signs of serious emotional distress.
OpenAI said it will monitor the rollout and share updates as the system develops. The work is informed by research into child development and advice from safety and medical organisations.
The company added that the age-estimation feature will begin rolling out in the European Union in the coming weeks to meet local regulatory requirements.
Also Read: OpenAI, ServiceNow Join Hands to Develop AI Agents for Business Workflows
