A Matter of Trust: Principles to Ethically Assess AI in Health Care.

Abstract

In this article, we focus on questions of agency in emerging technologies related to decision-making in medicine. We discuss three principles that were subsumed when bioethics embraced principlism: consent, confidentiality, and veracity. We argue that the advent of artificial intelligence and its employment within health care, impacts the physician-patient relationship in a way that its inclusion in other areas does not. In particular, we take up ethical dilemmas caused by AI related to trust, and illustrate how reflection on these subsumed principles helps to critique accurately policies related to the use of AI in health care and to navigate dilemmas associated with a loss of trust. We conclude by contrasting these principles with the proposed "five principles" system for AI, highlighting some areas of agreement, but also showing where consent, confidentiality, and veracity are necessary additions for ethically employing AI.