AI chatbots have failed people in crisis. Can that be fixed?
AI chatbots have repeatedly fallen short when users reach out in moments of crisis, according to clinicians and researchers who are now pressing AI companies to share the safety data that would reveal how often and how badly these failures occur.
Why this matters: People in crisis are among the most vulnerable users any technology will ever touch. When a chatbot gives a bad response at the wrong moment, the consequences are not a bad recommendation or a wasted afternoon. They can be serious and irreversible. Right now, the companies building these tools largely control what safety data gets seen and by whom. Clinicians and researchers cannot fix what they cannot measure. If AI companies want to deploy mental health features, they need to show their work.
Who should care: General readers · AI governance · Policy
This summary is AI-assisted and may contain errors. It is an original briefing to help you gauge significance quickly — not a reproduction of the source. Always read the linked original before relying on it. See our methodology.