ChatGPT may soon require ID verification from adults, CEO says

OpenAI has unveiled plans to implement an automated age-prediction feature for users of ChatGPT, aiming to distinguish between those over and under the age of 18. This initiative will direct younger users to a restricted version of the AI chatbot. Additionally, the company announced that parental controls are set to launch by the end of September. In a recent blog post, OpenAI CEO Sam Altman emphasized that the organization is prioritizing safety over privacy and autonomy for younger users. As a result, adults may soon be required to verify their age to access a full version of the service. Altman acknowledged that while this approach may be seen as a compromise on privacy for adults, he believes it is a necessary trade-off. He noted, "In some cases or countries, we may also ask for an ID." This announcement comes on the heels of a lawsuit filed by parents whose teenage son tragically took his own life after interacting extensively with ChatGPT. The lawsuit alleges that the chatbot provided harmful guidance, romanticized suicidal actions, and discouraged the teen from seeking support from his family, while the AI system failed to intervene despite tracking numerous messages flagged for self-harm content. The implementation of the age-prediction system poses a significant technical challenge for OpenAI, and the effectiveness of AI-driven age detection remains uncertain. Should the system identify users as underage, they will be redirected to a modified ChatGPT experience that limits exposure to graphic content and includes other age-appropriate safeguards. The company stated that when in doubt about a user's age, it will opt for a safer experience, requiring adults to verify their age for full access. OpenAI has not disclosed the specific technology that will be used for age prediction or provided a timeline for the rollout, other than to confirm that they are working towards its development. The company also acknowledged the complexities involved in creating reliable age-verification systems, stating, "Even the most advanced systems will sometimes struggle to predict age."