Perceptions of Justice By Algorithms

Abstract

Artificial Intelligence and algorithms are increasingly able to replace human workers in cognitively sophisticated tasks, including ones related to justice. Many governments and international organizations are discussing policies related to the application of algorithmic judges in courts. In this paper, we investigate the public perceptions of algorithmic judges. Across two experiments (N = 1,822), and an internal meta-analysis (N = 3,039), our results show that even though court users acknowledge several advantages of algorithms (i.e., cost and speed), they trust human judges more and have greater intentions to go to the court when a human (vs. an algorithmic) judge adjudicates. Additionally, we demonstrate that the extent that individuals trust algorithmic and human judges depends on the nature of the case: trust for algorithmic judges is especially low when legal cases involve emotional complexities (vs. technically complex or uncomplicated cases).

Algorithms in the court: does it matter which part of the judicial decision-making is automated?

“…The intermediate level of automation was preferred only in the first-information acquisition-stage. These results are consistent with the emerging literature on perceived algorithmic fairness within the law, which indicates that people might generally trust judges more than algorithms (Hermstrüwer and Langenbach 2022 ; Yalcin et al 2022 ). However, it also suggests that a binary view of judges vs. algorithms might be insufficient for capturing how people perceive automation in the court, as perceptions change depending on the stage of the decision-making process.…”

“…In particular, implementation, unlike analysis, might be perceived as an inherently human process, requiring capabilities that are simply irreplaceable by a computer (see, e.g., Kasy and Abebe 2021 ). This is particularly true if implementation involves emotions (Yalcin et al 2022 ; Xu 2022 ; Ranchordas 2022 ), e.g., allowing a human judge to incorporate equity concerns or compassion. Recall, however, that individuals with a legal profession in our study have an even stronger belief that automation in implementation is likely to yield unfair outcomes.…”

“…In the specific context of judicial decision making, recent studies show that people tend to trust human judges more than algorithms (see Yalcin et al 2022 ) or that they, at least, do not trust a fully-automated judicial process (see Hermstrüwer and Langenbach 2022 ). Kim and Phillips ( 2021 ) further argue that “a robot would need to earn its legitimacy as a moral regulator by demonstrating its capacities to make fair decisions”.…”

“…Conversely, their study finds that combining automated processes with high human involvement yields a higher fairness perception. Yalcin et al ( 2022 ) conducted a vignette experiment over MTurk and found that subjects care whether the judge is a human or an algorithm, finding evidence of higher trust in human judges.…”

From Bilateral to Ecosystemic Transparency: Aligning GDPR’s Transparency Obligations with the European Digital Ecosystem of Trust

“…Another line of research could follow-up on this and try to assess em pirically what individual recipients appreciate as satisfactory information sharing in case of an ADM, both in process and in content, and under which conditions information sharing results in actionable knowledge. This would certainly move beyond the current research into algorithm appreciation, 68 as it takes algorithmic decision-making as a given and focu ses more on contextual aspects. In that empirical project we would include the perceptions and practices of professionals that try to live up to the standards of transparency they are confronted with.…”