OpenAI is introducing an age-verification system in ChatGPT, aimed at improving safety for users under 18. The system, announced by CEO Sam Altman, includes an age-prediction tool that estimates a user’s age based on how they interact with ChatGPT. If the system is uncertain, it defaults to treating the user as a minor.

For users flagged as under 18, ChatGPT will offer a restricted version: graphic sexual content will be blocked, flirtatious responses curtailed, and conversations related to self-harm or suicide will be handled carefully, with emphasis on safety. 

OpenAI is also adding parental controls by month’s end. These features will allow parents to link their accounts with teen’s accounts, limit usage times or features (like chat history and memory), receive alerts if the teen shows signs of distress, and in rare cases contact authorities if there's a risk of imminent harm. 

This move follows a lawsuit by the family of a 16-year-old, Adam Raine, who died by suicide, alleging that prolonged interactions with ChatGPT contributed to his grief. 

OpenAI acknowledges that these changes involve trade-offs, especially around privacy, but says they are necessary to ensure teen users have a safer experience.