OpenAI has unveiled a new feature called Trusted Contact for its ChatGPT chatbot, designed to assist users at risk of self-harm. This feature allows individuals to nominate a friend who can be contacted if ChatGPT detects a serious risk of self-harm during conversations. The initiative comes after OpenAI faced scrutiny over the chatbot's previous interactions with users expressing suicidal thoughts, including a wrongful death lawsuit linked to a teenager's suicide. With Trusted Contact, users aged 18 and above can add a single adult as their Trusted Contact. This contact will receive a notification only after a trained team reviews the situation and determines that there is a significant risk. The notification will encourage the contact to check in on the user without sharing specific conversation details to maintain privacy. OpenAI emphasizes that while the system is not foolproof, it aims to enhance safety for users experiencing mental health crises.