Why Large Language Models Are Not Yet Safe for Autonomous Clinical Decision Support

A new perspective paper argues that even though LLMs can pass medical licensing exams, there's no real evidence they're safe to triage patients on their own, with no doctor in the loop.