Published Report Finds AI Model Developed for Teens Risky
A special chatbot version introduced by OpenAI for teens has been criticized due to shortcomings in safety measures and inadequacy during crisis moments.
OpenAI's AI version prepared for teens has become the target of criticism on the grounds that it fails to consistently apply safety measures and does not notify parents in crisis situations.
Shortcomings in Safety Measures
A new report published by Common Sense Media stated that the AI model specifically designed by OpenAI for teens contains an unacceptable level of risk. Researchers demanded that access be blocked for all teen users who do not have a verified adult account until the safety risks are resolved.
Parental Controls and Notification Issues
OpenAI defends that the new model is automatically activated for users between the ages of 13 and 17 and includes built-in controls that notify parents when dangerous topics such as suicide are brought up. However, the report found that no notifications were sent to parents during dialogues on sensitive topics such as suicide or eating disorders.
Study Mode and Guardrails
The model also features a study mode intended to guide students with their homework. Researchers stated that the guardrails of this mode can be disabled very easily and that the option to provide direct answers defeats the purpose of the feature.