The Ethical Dimension of AI-Systems in Healthcare: On Globalizing its Scope

Abstract

Artificial intelligence systems (AI-systems) are said to revolutionize healthcare. They are predicted to increase efficiency, free up time for more in-person care, lead to more accessible care, etc. However, these systems come with ethical reservations related to individuals’ privacy, dignity, social bias, etc. And although there are many approaches to the ethics of ai-systems in healthcare, most overlook these systems’ global impacts, such as their environmental and global social impacts. To meet this gap, I present some first steps to a global approach to the ethics of ai-systems in healthcare. In the first part of the presentation, I first describe some current and predicted uses of ai-systems in healthcare. Then, I shortly survey the most heavily discussed ethical issues related to these uses and present one particular common approach to meet these issues, the principlist approach. In the second part of the presentation, I argue that this common approach is unsuitable to meet the different environmental and the herewith related social impacts, such as increased health risks. Based on an interpretation of ai-systems as world-objects, which emphasizes the environmental and social materiality of ai-systems, I then present a global approach to the ethics of these systems which exists of five interrelated levels of ethical analysis and impacts: an individual-relational, organizational, societal, global, and historical level. The global approach to the ethics of ai-systems in healthcare complements current common ethical approaches, such as the principlist approach, by including ethical issues related to the environmental and social impacts, such as increased health risks, of ai-systems. Although there is already much ethical ground prepared to guide the use of ai-systems in healthcare, only by accounting for these systems’ global environmental and social impacts these systems can be ethically responsive for everyone and everywhere.