OpenAI expands ChatGPT parental alerts to cover violence threats

OpenAI is expanding parental alerts for ChatGPT. When a teen gets banned for violent threats or cyberbullying, parents who linked their accounts will now get a notice. Earlier, parents could already set time limits, reduce sensitive content, and get alerts if their child showed self-harm signs.

This new feature was built with Moonshot, a group that fights online violence. Its founder says telling parents about serious risks is a key first step to get help. The change comes after a 2025 shooting in Canada, where the suspect used ChatGPT. OpenAI didn't alert police then, and CEO Sam Altman later apologized.

There's also a new "learning mode" in parental controls. When on, ChatGPT won't give full answers right away but offers hints to help teens think. If teens use ChatGPT too long, they'll see more reminders to take breaks—helping build healthier habits.