An Attempt to Categorize Ethical and Social Issues of AI

Comparative analysis of issues through case studies

Abstract

It is very important to recognize all the potential risks arising ahead of the implementation of new AI technologies. However, it is very difficult to imagine the potential ethical and social issues of individual AI product in real society, in advance, in a comprehensive and precise manner. The conclusions will also be greatly influenced by individual ethical and social insights. Therefore, there is a need for an effective and reproducible method to consider ethical and social issues regarding such specific AI products. One such measure is an analogy-based approach based on the case studies, which has already been discussed as ethical and social issues in history. In this presentation, we introduce AI use cases in the following four areas: (1) AI in medical practice, (2) social AI, (3) crime prediction and AI, and (4) competitive games and AI, and discuss the existence of analogies between problems and the possibility of categorization by comparing case studies between domains.