The Morality of Artificial Agents

Abstract

The informational approach has powerful practical and theoretical consequences on the notions of agency and responsibility. This chapter discusses one of the most crucial questions facing our networked information society today: who is the “who” that performs an action and may be held accountable for the (moral or legal) consequences of that action? It is increasingly likely that the agent is an artificial one, whose autonomous performance of tasks poses the risk of unexpected and potentially harmful effects. How are we to deal with consequences that cannot be traced back directly to a human being? Here we explore insights that IE might offer into the questions at hand. Investigation of the notions of agency and autonomy from an informational point of view is useful in shedding light on the moral experience of accountability and responsibility. As with patients of the moral experience, by extending the class of moral agents in the direction of non-human agents, we can better grasp the role of humans in moral (as well as in legal) responsibility, and are better equipped to deal with the novel theoretical and practical problems arising from our interaction with artificial autonomous agents.