Sentencing and Algorithmic Transparency

Algorithmic transparency has received much attention in the current debate on the use of AI in sentencing. Concern has been expressed that the introduction or increased use of algorithms in the sentencing process may challenge the scrutability of and insight into the decision-making. However, there is a striking contrast between the repeated emphasis on the alleged challenge of algorithmic opacity and the lack of elaborate attempts at answering precisely what this challenge consists of. Drawing on the traditional discussion of the importance of judicial reason-giving, this chapter examines various reasons as to why and with regard to what algorithmic transparency may be morally important. It is argued that adherence to openness in the form of judicial reason-giving does not necessarily commit one to a view with regard to algorithmic transparency. Furthermore, even though there are strong reasons in favor of algorithmic transparency, the field still suffers from a lack of empirical research.