ChatGPT Introduces ‘Trusted Contact’ Feature to Notify Loved Ones in Case of Safety Issues

OpenAI is rolling out a new optional safety feature for ChatGPT that allows adult users to designate an emergency contact for mental health and safety issues. This “Trusted Contact” can be a friend, family member, or caregiver who will be alerted if OpenAI’s system identifies discussions related to self-harm or suicide.

The concept behind the Trusted Contact feature is straightforward and backed by expert validation: when someone is in crisis, reaching out to someone they trust can significantly help. This feature adds an extra layer of support alongside the local helplines already accessible in ChatGPT.

Participation in the Trusted Contact feature is voluntary. Any adult user can activate it by entering the contact information of another adult (18+ globally or 19+ in South Korea) in their account settings. The designated Trusted Contact must accept the invitation within a week. Users have the flexibility to modify or remove their contact at any time, and the Trusted Contact can also opt out whenever they choose.

OpenAI emphasizes that the notifications are “intentionally limited” and will not disclose chat details or transcripts to the Trusted Contact. If the system detects that a user is discussing self-harm, ChatGPT will prompt them to reach out to their Trusted Contact for assistance, informing them that the contact may be notified. A specialized team will then assess the situation, and if serious safety concerns are identified, ChatGPT will send a brief email, text, or in-app notification to the Trusted Contact.

This feature builds on the emergency contact option introduced with ChatGPT’s parental controls in September, following a tragic incident involving a 16-year-old who took his own life after confiding in the chatbot. Similarly, Meta has implemented a feature that notifies parents if their children frequently search for self-harm topics on Instagram.