Addressing User Resistance Would Have Prevented a Healthcare AI Project Failure

Abstract

Integrating artificial intelligence (AI) into existing work routines involves invasive changes, and the resulting user resistance can lead to project failure. We describe a failed AI project at a large hospital to implement a cognitive agent and identify the root causes of the user resistance that led to the failure. Based on the lessons learned, we provide recommendations for addressing the causes of resistance for the three types of AI-automation, decision support and engagement.