ChatGPT's New Safety Tool Lets Users Add a Trusted Contact for Mental Health Support

ChatGPT has introduced a new safety feature allowing adult users to designate a 'Trusted Contact' in case of potential mental health crises. The tool aims to provide support while maintaining privacy, with notifications sent only when serious concerns arise. Users aged 18 and over—or 19 in South Korea—can now add one adult contact through their account settings.

The feature works by automatically monitoring conversations for language that may indicate self-harm. If such content is detected, ChatGPT first alerts the user that their Trusted Contact might be notified. This prompt encourages direct communication before any action is taken.

When concerning language is flagged, a dedicated team reviews the situation. OpenAI has committed to assessing these cases in under one hour, with every notification undergoing trained human review. If a serious risk is confirmed, a limited alert is sent to the Trusted Contact—sharing only that self-harm was discussed and suggesting a check-in. No chat details are disclosed to protect user privacy. Development of the tool involved input from OpenAI’s Global Physicians Network and its Expert Council on Well-Being and AI. Experts like Dr. Arthur Evans, CEO of the American Psychological Association, highlight the role of social connection in mental health crises. Meanwhile, Dr. Munmun De Choudhury, a professor at Georgia Tech, notes that AI can help foster genuine human support and psychological safety.

The Trusted Contact feature combines automated monitoring with human oversight to address potential self-harm risks. Users retain control over who is notified, and privacy remains a priority throughout the process. OpenAI’s approach reflects a broader effort to integrate safety measures into AI-driven interactions.