ChatGPT: OpenAI Launches 'Trusted Contact' for Safety

OpenAI recently unveiled a new safety feature for ChatGPT, called 'Trusted Contact'. This option, aimed at adult users, allows them to designate a close person who will be alerted in case of concerns related to mental health or safety. This mechanism comes into play when the system detects discussions around self-harm or suicide.

The feature is based on a principle validated by experts: in times of crisis, the support of a trusted person can be crucial. OpenAI emphasizes that this adds an extra layer of support to the local helplines already integrated into ChatGPT.

Activating 'Trusted Contact' is optional. Users can add the contact information of an adult (18 years and older, or 19 years in South Korea) in their account settings. The designated contact must accept the invitation within a week. Users have the option to modify or remove this contact at any time, and the contact can also choose to opt out.

OpenAI specifies that notifications sent to the 'Trusted Contact' are intentionally limited. They do not contain details about the conversations or transcripts. If automated systems detect concerning discussions, ChatGPT encourages the user to reach out to their 'Trusted Contact'. A small trained team then reviews the situation before sending a brief notification via email, text message, or through the ChatGPT app.

This initiative is part of the ongoing safety measures introduced with ChatGPT's parental controls in September, following a tragic incident involving a teenager. Meta has also implemented a similar feature on Instagram, alerting parents when their children repeatedly search for self-harm topics.