ChatGPT's Trusted Contacts Feature Supports Users in Mental Health Crises
In May 2024, OpenAI introduced a new Trusted Contacts feature within ChatGPT to help users struggling with mental health issues connect with close contacts. Users can proactively register one or more trusted individuals so that if ChatGPT detects signs of suicide risk or self-harm in conversations, it automatically notifies those contacts to provide timely assistance.
Designed to prevent suicide and support users during emotional crises, Trusted Contacts employs AI algorithms to identify potential mental health concerns based on chat content. Upon detecting severe risk, the system reaches out to pre-registered contacts, creating a rapid response channel for urgent help.
However, this feature only activates if users have manually set it up in their ChatGPT account settings; the chatbot cannot enable or configure it on their behalf. Without prior registration of trusted contacts, ChatGPT cannot send emergency alerts. Additionally, the feature is currently limited to text input and is not available in voice-enabled chatbot versions.
Real-World Experience with Feature Visibility and User Onboarding
NPR reporters found that ChatGPT does not proactively introduce or remind new users about the Trusted Contacts feature during initial smartphone app setup and first interactions. The chatbot greets users and inquires about topics of interest but omits mention of this mental health support tool.











