A Brief Overview of an Approach Towards Ethical Decision-Making

Abstract

Ethics in decision-making reflects traits such as transparency, equity, and trust. However, when considering ethics in the decision-making process of autonomous agents, the significant challenge is how autonomous agents should interact to reach an agreement, knowing that their ethical preferences may differ. On that account, this study explores two fields to propose an approach to ethical decision-making: automated negotiation, the field concerning interaction among multiple agents to reach an agreement, and machine ethics, the field concerned with adding or ensuring moral behaviors from agents. Although agents can negotiate and decide on a solution automatically, whether they can propose an ethically correct decision is still a subject matter. To this end, this study proposes the concept of introducing ethics in the decision-making process of intelligent agents for ethical decision-making. In particular, we propose a research framework that addresses how user ethical preferences can be converted into quantifiable measures and further used by autonomous agents during negotiation for ethical decision-making.