OpenAI has updated its internal Model Spec for ChatGPT to place the safety of teenage users above other system goals, marking a shift in how the artificial intelligence platform is designed to respond to users aged 13 to 17. The change reflects growing scrutiny around the impact of generative AI tools on minors.
Under the revised guidelines, protections for teenagers are given precedence even when they may limit the model’s ability to maximise helpfulness or autonomy. OpenAI said the update is intended to ensure age-appropriate interactions and reduce potential harm in sensitive or high-risk conversations involving younger users.
The revised framework introduces a set of Under-18 principles that guide how the model should behave when interacting with teens. These principles are informed by developmental research and expert inputs, and focus on prevention, transparency, and early intervention. The model is expected to encourage safer alternatives, promote responsible use, and guide users towards trusted real-world support when required.
OpenAI said the updated safeguards apply particularly to topics such as self-harm, suicide, explicit content, substance use, dangerous activities, and unhealthy behavioural advice. In such cases, the system is designed to avoid reinforcing harmful behaviour and, where appropriate, suggest reaching out to family members, trusted adults, or professional support services.
As part of the changes, OpenAI has also begun early testing of an age prediction system that can infer a user’s likely age based on conversational signals. If a user’s age cannot be clearly determined, the platform defaults to the under-18 experience, applying stricter safety controls. Adult users may be offered options to verify their age to access the standard experience.
The update comes amid increasing regulatory attention globally, as policymakers and child safety advocates push for stronger safeguards for minors on digital platforms. OpenAI said the revised model guidelines are part of its broader efforts to align product design with emerging safety expectations and regulatory standards.
Keep building. Keep learning. Keep growing with StartupByDoc.