Exploring Physicians’ Experiences of Conflicts Between Clinical Judgment and AI Recommendations in the Use of AI-Based Clinical Decision Systems

Abstract

The integration of Artificial Intelligence (AI) in healthcare has transformed clinical decision-making, raising important questions about how medical professionals adapt to and experience this technological shift. While prior studies have explored system performance and usability, little is known about how physicians subjectively interpret their interactions with AI-based decision support systems in real-world settings. This study addresses the gap by asking: how do physicians experience conflicts between their clinical judgment and AI-generated recommendations? Using an interpretative phenomenological approach, this study explores the lived experiences of physicians navigating such conflicts within tertiary hospital environments in Indonesia. In-depth, semi-structured interviews were conducted with ten physicians from three tertiary-care hospitals, all of whom had at least six months of experience using AI-supported diagnostic systems. Data were analyzed using interpretative phenomenological analysis (IPA) to uncover key themes related to emotional tension, ethical ambiguity, and shifting professional identity. The findings reveal that physicians perceive AI not merely as a tool, but as an influential presence that challenges their sense of clinical autonomy and accountability. Participants also described adaptive strategies for reconciling AI recommendations with their professional judgment over time. These results deepen our understanding of the ethical and experiential dimensions of AI integration in clinical practice, especially in low- and middle-resource healthcare systems such as those found in Southeast Asia. The study highlights the need for more human-centered AI implementation strategies and calls for continued exploration of clinician experiences as technology reshapes medical practice.