Proceedings of the Annual Meeting of the Cognitive Science Society

Moral judgments are elicited using dilemmas presentinghypothetical situations in which an agent must choosebetween letting several people die or sacrificing one person inorder to save them. The evaluation of the action or inaction ofa human agent is compared to those of two artificial agents –a humanoid robot and an automated system. Ratings ofrightness, blamefulness and moral permissibility of action orinaction in incidental and instrumental moral dilemmas areused. The results show that for the artificial cognitive agentsthe utilitarian action is rated as more morally permissible thaninaction. The humanoid robot is found to be less blameworthyfor his choices compared to the human agent or to theautomated system. Action is found to be more appropriate,morally permissible, more right, and less blameworthy thaninaction only for the incidental scenarios. The results areinterpreted and discussed from the perspective of perceivedmoral agency.