Safety Organization Flags ChatGPT Teen Version for Failing to Alert Parents
Common Sense Media released an assessment criticizing OpenAI's ChatGPT for Teens, claiming its safety features fall short of promises made to parents and users. The nonprofit found that the service fails to send appropriate alerts to guardians about concerning conversations, does not adequately respond to crisis situations, and still completes homework assignments despite being designed for educational use. OpenAI disputed the findings, stating its commitment to teen safety and willingness to undergo independent testing.
Common Sense Media's Youth AI Safety Institute conducted testing of OpenAI's teen-specific ChatGPT version and documented multiple concerning gaps in its protective mechanisms. The organization's evaluation revealed that the system failed to notify parents about extended conversations involving self-harm topics, provided inadequate support during mental health emergencies, and continued assisting with academic work despite educational positioning. OpenAI contested these findings, suggesting that testing occurred before parental control systems were fully activated, while Common Sense Media maintained that their methodology was sound and that notification failures persisted even in accounts established well after activation periods.
This dispute highlights ongoing tensions between AI companies and child safety advocates regarding product transparency and accountability. Parents relying on these safeguards may face genuine risks if protection features underperform, while broader questions emerge about how technology firms validate safety claims before public release. The disagreement over testing methodology could influence regulatory approaches to youth-focused digital tools and set precedent for how independent safety assessments are conducted in the AI industry.