Our commitment to community safety

OpenAI outlines its multi-layered approach to preventing the misuse of ChatGPT for violence, harm, or illegal activities. The company employs model training, automated detection systems, and expert-led human reviews to identify risks while maintaining support for educational or historical inquiries. Safety measures include flagging concerning patterns in long-term conversations, providing crisis resources for self-harm, and implementing strict enforcement policies, including account bans and notifications to law enforcement in cases of imminent, credible threats. Additionally, OpenAI is expanding safety features like parental controls and trusted contact options to further protect users.

(Source：OpenAI)