OpenAI's ChatGPT: New Trusted Contact for Self-Harm Alerts

OpenAI has launched a new "Trusted Contact" feature for ChatGPT users. Here's the thing: adult users can now pick a contact who gets an alert if ChatGPT detects conversations about potential self-harm. This builds on parental controls introduced in September 2025. To set it up, you add one adult as your Trusted Contact in the settings. They get an invitation and have a week to accept. If they decline, you can choose someone else. What's interesting is that when the system flags a conversation, ChatGPT first tells the user it might notify their contact. Then, human reviewers confirm the seriousness before a brief alert is sent via email, text, or in-app notification. OpenAI worked with mental health experts to develop this, highlighting the importance of social connections during distress. This feature aims to provide real-world support for users in crisis.