Conceptual engineering for ethics of social robots: on fictionality of moral agency

Abstract

Moral agency is one of the most commonly discussed concepts in robot ethics. In this paper, we consider the moral agency of social robots in terms of fiction theory. As taking the user's emotional responses into consideration is important in social robotics, some researchers have developed a "relational turn" approach to robot ethics. Coeckelbergh, for example, has discussed the moral significance of robots in terms of "virtual moral agency." However, such an extension of moral agency creates confusion in whether the moral responsibility is attributed to human or robots. Therefore, we suggest an alternative approach that distinguishes between social robots' human likeness and moral status. Specifically, based on Rodogno's argument, we try to understand social robots as "props" that create a fictional world. We believe that this approach provides an innovative solution to explaining the moral status of social robots without extending the concept of moral agency recklessly.