ChatGPT's Safety Net: Trusted Contact Feature?

An optional feature within ChatGPT, called Trusted Contact, aims to connect users contemplating suicide or self-harm with a person they trust during emotional crises. This feature was developed by OpenAI after it was estimated that over a million weekly users discussed such topics. The chatbot is designed to detect serious safety risks and alert a pre-designated Trusted Contact. This feature, however, is not actively promoted within the app, and users must opt-in by setting it up in their account settings prior to any crisis. Consequently, if a user is in severe distress and has not set up the feature, ChatGPT cannot notify their contact. Mental health experts have offered varied opinions on the feature's potential. Some, like Holly Wilcox, believe it would be ideal if offered proactively to all users. Ursula Whiteside notes that while the setup is simple, its effectiveness in a crisis depends on the user's state of mind. Jennifer Lockman raises concerns that a chatbot might not overcome an individual's internal barriers to seeking help, emphasizing that it cannot replicate the nuanced guidance of a mental health professional. Despite these reservations, Lockman acknowledges it as another avenue for promoting human support and connection, a critical protective factor against suicide. OpenAI states it is collaborating with experts to refine when and how the feature is offered, alongside other safeguards like crisis hotline information.