Abstract

The purpose of Machine Ethics is to develop autonomous technologies that are able to manage not just the technical aspects of a tasks, but also the ethical ones. As a consequence, the notion of Artificial Moral Agent (AMA) has become a fundamental element of the discussion. Its meaning, however, remains rather unclear. Depending on the author or the context, the same expression stands for essentially different concepts. This casts a suspicious light on the philosophical significance of Machine Ethics. In particular, the risk arises of discarding Machine Ethics as a whole on the basis of accusations that, however, apply exclusively to one specific understanding of what AMAs are – but not to other, more adequate and convincing conceptualisations. To avoid this pitfall, this essay tries to elaborate a philosophically sound interpretation of AMAs and to sketch its primary component, i.e., the notion of functional ethics.