Machine-Mediated Morality: Do People Want Robots to Be More Utilitarian or Less Omissive than Humans?

Abstract

Abstract Experimental studies on preferences towards human and robot agents in instrumental dilemmas suggest that people have higher utilitarian preferences for machines than for humans. However, these studies give participants only two options for the agent: either a utilitarian action or omission. It, thus, remains unclear whether a higher share of utilitarian preferences means that people become more utilitarian when considering robot agents or whether they just grant a human more leeway for omission. This study aims to disentangle these effects in two ways: first, by offering an alternative actionable choice, and second, by connecting the utilitarian outcome with an omission. Results suggest that people become, in fact, more utilitarian-minded, when considering their moral preferences for machines.