ChatGPT Teen Version Fails Safety Tests, Report Warns

Independent researchers have found that ChatGPT for Teens, OpenAI's variant designed specifically for younger users, contains significant safety vulnerabilities that could expose children to harm. The study revealed that the platform's built-in safeguards fail to consistently identify when young users are experiencing mental health emergencies or in psychological distress.

The investigation highlighted multiple concerning gaps in the application's protective mechanisms. Most notably, the system does not reliably alert parents or guardians when teenagers encounter potentially harmful content or express warning signs related to mental wellbeing. This represents a substantial oversight for a platform marketed with enhanced protections for minors.

Additionally, researchers discovered that the safety features can be circumvented relatively easily, allowing students to bypass learning restrictions that were intended to prevent the tool from replacing educational instruction. The study suggests these workarounds undermine the platform's core safeguarding objectives.

The findings come at a time of increased scrutiny regarding artificial intelligence applications targeting young people. Parents and educators across Asia have grown increasingly concerned about how AI tools affect children's mental health, learning outcomes, and online safety. OpenAI has not yet publicly responded to the study's conclusions.

Experts recommend that technology companies implementing AI platforms for minors conduct more rigorous safety evaluations before launch and maintain ongoing monitoring systems to address emerging vulnerabilities.