Report: ChatGPT for Teens failed to notify parents in tests
Technology
ChatGPT for Teens, OpenAI's AI chatbot for ages 13 to 17, is under fire after a report said it doesn't alert parents about serious topics like self-harm or suicide.
In tests, explicit self-harm, suicide, and eating-disorder prompts never triggered a parental notification, and only four parental notifications were triggered on older accounts used over weeks.
OpenAI rebuts review, feature concerns remain
OpenAI pushed back on the report, saying its methods were flawed and that teen safety is a top priority.
Still, the review also pointed out that features like "Study Mode" can be turned off easily by teens and the age-prediction tool isn't reliable.
Common Sense Media says big improvements are needed before parents can really trust this AI with their kids' well-being.