Some guardrails on ChatGPT for Teens don't work as promised, watchdog group says

A watchdog group has criticized OpenAI's ChatGPT for Teens product, stating that some of its intended safeguards do not function as promised and that the chatbot poses an "unacceptable risk" to children. Common Sense Media, the group, called on OpenAI to restrict the chatbot to adults until it provides a safe and developmentally appropriate experience.

OpenAI refuted the report's findings, claiming they may not have reflected the effectiveness of OpenAI's safeguards. The report noted that while some teen-mode features, such as the chatbot's refusal of explicit sexual role-play, worked as intended, others, like alerts meant to notify parents of suicide, self-harm, or eating disorders risks, did not.

Common Sense Media stated that ChatGPT still behaves like a friend to teenagers, despite OpenAI's promise to prevent such interactions. OpenAI emphasized its commitment to teen safety and developing safeguards, adding that its review of the report's methodology suggests inaccuracies in Common Sense Media's findings.