The Ethics of Empathetic AI in Medicine

Abstract

The expression of empathy is an important part of effective and humane medical care. Modern medicine faces a significant challenge in this area, at least in part due to the ever-increasing demands on clinicians’ time. When empathic responses to patient emotions are lacking, the quality of care suffers, including a patient’s ability to understand important information and willingness to trust their doctor. This article asks whether AI can help address this challenge. We approach the question by presenting and critically analyzing three novel scenarios that represent fundamentally different approaches, including (1) AI-powered “doctor” avatars directly interacting with patients, (2) AI editors and co-authors helping clinicians find the right words, and (3) AI simulated patients helping to provide communication training. In each case, we identify fundamental AI and medical ethics considerations, including those that warrant further consideration.