A Framework for Understanding the Ethics of Human-AI Interactions. Exploring the Moral Dimensions of Artificial Intelligence Entities

As artificial intelligence entities, designed to be Learning Machines, gain cognitive abilities and autonomy, they lose their ethical neutrality: they become moral entities, which must be trained to exhibit socially acceptable behaviors. In such a largely unexpected context, this paper proposes a framework to improve our understanding of the challenges arising in interactions between humans and intelligent artificial entities, systematically classifying these issues according to a well-grounded theoretical scheme, and assessing them from a moral perspective attuned to a rapidly evolving scenario. This framework offers insights into the requirements for ethically sensitive management of human-AI interactions and highlights some challenges already emerging from the diffusion of autonomous Learning Machines.