A cautious defense of AI as moral machines

Abstract

Developments in the field of Artificial Intelligence have called into question whether it is defensible to allow AI to make and execute some of our moral decisions. This paper explores some of the motivations and challenges in doing so, and attempts to provide a well-balanced and comprehensive approach to the development and implementation of AI. I argue that while we may encounter serious difficulties, further work can help us formulate a competent ethical framework of AI such that we can benefit from AI’s capacity for good.