This chapter discusses the ethical role of explainability in Artificial Intelligence (AI), addressing the question of whether and how it would be morally relevant to understand the functioning of an algorithm. A discussion is made on the conceptualization of different terms, such as explainability, explicability, and interpretability. Currently, there is no consensus on the technical definition of explainability, complicating discussions about its ethical implications; even so, some characteristics attributed to it allow the debate. The chapter argues that the “black box” nature of some Machine Learning algorithms presents a qualitatively new ethical problem, as their outputs can be, in some sense, incomprehensible even to their developers, unlike other technologies. This chapter also contests the proposal to establish “Explicability” as a fifth ethical principle in AI on par with the four principles of bioethics’ Principlism (Respect for Autonomy, Nonmaleficence, Beneficence, and Justice), arguing instead that Explainability has an instrumental value, being morally relevant for AI but not a principle in itself.