Users aged 13 to 17 will automatically be placed into a restricted version of ChatGPT that prohibits the AI from using romantic language or claiming it has consciousness. The rollout began August 18 for free and paid personal accounts globally, with full availability in Australia on September 8. The system includes a "Study Mode" that redirects shortcut attempts to learning tools and provides parental controls for scheduling quiet hours and receiving safety notifications regarding eating disorders.
The update follows lawsuits alleging that inappropriate chatbot conversations contributed to the harm or deaths of young people. OpenAI updated its Model Spec to prevent the AI from encouraging emotional dependence and announced a partnership with CodeAI for student and educator training. Safety protections are enabled by default and cover self-harm, violence, and explicit sexual content.
Key sources
- SOURCE@cnn“lawsuits alleging that inappropriate conversations with ChatGPT contributed to the harm or deaths of young people”x.com
- SUPPORT@wallstengine“If ChatGPT estimates a user is under 18, or they say they’re 13–17, they’ll automatically enter it”x.com
- SUPPORT@btibor91“full availability in Australia expected September 8”x.com
- SUPPORT@kimmonismus“built to act less like an answer machine, and less like a person”x.com
- SUPPORT@ft“limits teens to dedicated version of ChatGPT promising more safeguards”x.com