Artificial intelligence and the erosion of professional autonomy in clinical practice – a review and analysis

Professional autonomy is a foundational ethical principle in healthcare, underpinning healthcare professionals’ capacity to exercise independent clinical judgement in light of a patient's particular circumstances, preferences and needs. While ethical debate surrounding medical AI has focused extensively on safety, accuracy and bias, its implications for professional autonomy have received comparatively less attention, a notable omission given AI's growing role in shaping clinical decision-making. This article identifies three salient mechanisms through which AI integration may constrain professional autonomy: (1) algorithmic opacity and explainability, (2) automation bias and de-skilling and (3) institutional pressures. Whilst proposed safeguards such as explainable AI, professional training, system design and institutional governance represent important directions, they do not yet adequately protect professional autonomy. This article therefore calls for closer scrutiny of current approaches to AI integration and for governance frameworks that attend not only to the technical properties of AI systems, but to the organisational and institutional conditions in which they are deployed.