OpenAI adds parental controls to ChatGPT: How they work
What's the story
OpenAI has launched new parental controls for its AI chatbot, ChatGPT. The updated features allow parents to link their accounts with their children's and set restrictions like quiet hours and content filters. The changes also include stronger protections against graphic content and a self-harm notification system.
Enhanced controls
Parents can set limits like quiet hours, disable voice mode
The new update, which is available from today, lets parents link their account with their teen's. This way, they can set limits like quiet hours, disable voice mode and image generation, prevent memory usage and model training. However, these changes can't be made by teens themselves. Only parents have the authority to adjust settings as required.
Protection measures
Teens get stronger protections against graphic content
Along with the controls, teen accounts linked to parents will get stronger safeguards automatically. These include reduced exposure to graphic content, viral challenges, romantic or violent roleplay and extreme beauty ideals. While parents can choose to disable these protections if they wish, teens can't do it on their own.
Alert feature
Self-harm detection and parent notification system
OpenAI has also added a notification system to alert parents if ChatGPT detects signs that their teen may be considering self-harm. In such cases, trained staff will review the situation and if there's evidence of acute distress, parents will be notified via email, text or phone alerts. In extreme cases, law enforcement or emergency services could get involved in case of an imminent threat.
Future plans
Parent resource page launched
OpenAI has also launched a parent resource page with information on how ChatGPT works, guides for setting up controls, and suggestions for positive use in schoolwork, creativity, and family activities. The company plans to expand this resource page with expert advice and conversation tips over time. OpenAI is also developing an age prediction system that will automatically apply teen-appropriate settings if a user is under 18.