Implementing artificial intelligence ethics: A tutorial

Abstract

Artificial Intelligence (AI) Ethics have steadily gained attention following various real-world incidents surrounding both purely digital and cyber-physical AI systems. Concerns have been raised over the ethical aspects of these systems for example in relation to data privacy, or material harm in the case of cyber-physical systems. Though academic activity in the area has grown recently, much of the current corpus consists of theoretical and conceptual studies. Attempts to bring this on-going discussion into practice have been primarily made in the form of various guidelines for ethical development of AI systems. However, these guidelines have not been adopted out on the field. The current situation in the area calls for more actionable methods for AI ethics, focusing on the point of view of developers. In this paper, we discuss current methods for implementing ethics in different contexts and then provide an introduction to a tutorial on a developer-focused method for implementing AI ethics, the Ethics Card Deck.