Overview
- China’s cyberspace regulator released a draft for public comment that would require age verification and guardian consent for minors using human‑like chatbots.
- Proposed Chinese safeguards include bans on content tied to suicide, self‑harm, gambling, obscenity and violence, time limits for minors, and mandatory human takeover with guardian alerts when suicide risk appears.
- Australia has already activated enforceable rules preventing AI services from serving under‑18s pornographic material, sexually explicit content, self‑harm or disordered‑eating content, according to the eSafety Commissioner.
- OpenAI updated its Model Spec and teen tools for users 13–17, restricting romantic or sexual roleplay and violent roleplay, tightening guidance on body‑image topics, and emphasizing real‑time risk detection and parental controls.
- Surveys and cases underscore urgency, with studies reporting widespread teen use of AI companions for sensitive topics and ongoing lawsuits alleging links between chatbot interactions and self‑harm, as experts flag unresolved challenges in age checks, real‑time escalation and independent oversight.