Clinicians and researchers urge AI firms to release safety data after chatbot crisis failures

A collective of clinicians and research experts has strongly asserted that artificial intelligence companies must transparently share their safety data with the broader scientific and medical communities. This call to action follows mounting evidence and growing concerns that current conversational AI models have fallen short when tasked with assisting users who are navigating severe emotional or psychological crises.

This issue has sparked widespread debate as large language models and chatbots become deeply integrated into daily routines, leading some vulnerable individuals to seek guidance from automated systems during critical moments. Unfortunately, the current infrastructure lacks the rigorous safety standards and specialized protocols necessary to properly assess and de-escalate crisis situations in real time.