Artificial Intelligence and SentencingHumans against Machines

Abstract

Abstract In this chapter, we explore whether AI would be better in making legitimate sentencing decisions than human judges. In order to do so, we first develop a multilayered model to assess the legitimacy of sentencing, and then compare human and machine judges (current machine learning AI and futuristic AI as a full moral agent) on their abilities to achieve legitimate sentencing. We conclude that human judges and machine judges have their own strengths and weaknesses in different aspects of our model. However, we believe the “human” in sentencing is so important, that even futuristic super AI should never be entrusted to fully take over sentencing from human judges.