OpenAI has introduced its new U18 Principles, aimed at enhancing the safety of teenage users interacting with ChatGPT. This update, part of the Model Specification revisions, focuses on protecting individuals aged 13 to 17 and is designed in collaboration with the American Psychological Association (APA).
The new guidelines were announced in December 2025, following the previous Teen Safety Blueprint released in November of the same year. This initiative comes amid increasing scrutiny from regulators and public concern regarding the safety of minors online.
The U18 Principles are built upon four foundational commitments. These include prioritizing the safety of teens, fostering real-world support and offline connections, appropriately addressing teenagers without treating them as children or adults, and ensuring transparency regarding the AI”s limitations. OpenAI”s approach will involve exercising heightened caution when discussions with teen users touch on sensitive issues such as self-harm, explicit content, and other potentially harmful behaviors.
Dr. Arthur C. Evans Jr., CEO of the APA, emphasized the importance of providing developmentally suitable protections for young users, highlighting that human interaction is crucial during adolescent development. He advocates for a balanced use of AI alongside real-world relationships.
To implement these principles, OpenAI is deploying an age-prediction model across its consumer ChatGPT services. This system automatically activates teen protections when it identifies a user as a minor. For accounts with ambiguous or incomplete age information, the U18 experience will be the default setting until the user”s adult status is confirmed.
Additionally, the enhancements extend parental controls to newer features such as group chats and the ChatGPT Atlas browser. OpenAI has also partnered with ThroughLine to integrate localized crisis hotlines within ChatGPT and its Sora platform, ensuring users have access to necessary support in real-time.
This update aligns with broader industry trends as major tech firms increasingly focus on the safety of younger audiences. OpenAI”s establishment of an Expert Council on Well-being and AI in October 2025 signifies its commitment to guiding the development of healthy AI interactions across various age demographics. A global network of medical professionals is involved in assessing the AI”s conduct in sensitive discussions.
For investors, the development of this teen safety framework represents both a compliance measure and a competitive advantage. Companies that proactively implement robust protections for youth may encounter fewer regulatory challenges as global oversight of AI technologies intensifies. OpenAI has pledged to continue refining its principles in response to emerging research, expert insights, and practical application data.
