Abstract: Philosophical research in AI has hitherto largely focused on the ethics of AI. Here we suggest that the time has come to start a novel area of research in AI – the epistemology of AI, and in particular the ethics of belief of AI. This paper is a collaboration between an ethicist of belief and a machine learning engineer (Dr. Vincent Valton) to investigate the ways in which extant work in the ethics of (human) belief can be applied to an ethics of AI belief, and in exciting new topics of interest in the latter domain. The focus of today’s talk will be on the ethics of AI belief with respect to AI profiling beliefs and algorithmic discrimination. I here understand “profiling” as forming a belief about an individual on the basis of well-evidenced (by which I don’t mean morally justified) generalizations about a group or groups to which that individual belongs. And I understand “discrimination” as the differential treatment of an individual based on their membership in a particular group or groups.

(Joint Work-in-Progress by Winnie Ma and Vincent Valton)