OpenAI says teen ChatGPT use limited but research finds it an 'unacceptable risk'

OpenAI, the artificial intelligence company, initially launched ChatGPT specifically for teens in August to encourage healthier usage of the AI chatbot. The new safety features included strict regulations against emotional dependence, parental notifications for problematic usage, and restrictions on sexualized content. However, a recent study from the non-profit Common Sense Media has revealed that despite these measures, ChatGPT for Teens still poses a significant risk for teenagers.

The research found that even though average teen use of ChatGPT was limited, often limited to under 15 minutes a day, the safety features, including parental alerts for self-harm conversations, were often ineffective. Even a single hour of conversation about suicidal thoughts, self-harm or disordered eating, triggered no alerts to parents.

The report also noted that the platform failed to recommend professional help to teens in crisis. Despite OpenAI's assurances to curb inappropriate behavior, the bot continued to engage in anthropomorphized language, giving parents a false sense of security. This comes at a time when concerns about online safety and mental health among children are on the rise, with several lawsuits being filed against tech giants such as Meta and TikTok.

OpenAI, too, is under increased scrutiny and facing legal pressures over its handling of young users in ChatGPT.