Artificial intelligence in healthcare: Doctors, patients and liabilities

Abstract

AI is increasingly finding its way into medical research and everyday healthcare. However, the clear benefits offered to patients are accompanied not only by general limitations typical of the application of AI systems but also by challenges that specifically characterize the operationalization of the concepts of disease and health. Traditionally, these challenges have been dealt with in the physician-patient relationship in both medical ethics and civil law. The potential for incorrect decisions (and the question of who is responsible for such decisions) in cases where AI is used in a medical context calls for a differentiated implementation of medical ethical principles and a graduated model of liability law. Nevertheless, on closer examination of both fields covering relevant obligations towards patients and users against the backdrop of current medical use cases of AI, it seems that despite a certain level of differentiation in the assignment of responsibilities through rules on liability, those affected, in the end, are generally left to deal with any AI-specific risks and damages on their own. The role played by the physician in all this remains unclear. Taking into account the physician-patient relationship as a contractual obligation in a broad sense can assist in clarifying physicians’ roles and determining their duties in a sustainable and patient-friendly manner when applying AI-based medical systems. This can contribute to reinforcing their established ethical and legal status in the context of AI applications.