Empirical Investigation of Algorithmic and Judicial Bias in Online Dispute Resolution Proceedings

Abstract

This study aims to empirically investigate the manifestation of algorithmic bias and judicial bias within online dispute resolution (ODR) systems, particularly in the context of Artificial Intelligence (AI)-assisted decision-making. Employing a qualitative case study approach, the research explores how bias is constructed, experienced, and negotiated by various actors involved in digital dispute resolution. Data were collected through semi-structured interviews, limited observations, and document analysis involving judges, mediators, legal practitioners, system developers, and ODR users selected through purposive and snowball sampling techniques. The findings reveal several key themes, including algorithmic opacity leading to epistemic asymmetry, the persistence and transformation of judicial bias within hybrid human–AI decision-making processes, and the relational construction of trust and legitimacy intertwined with the reproduction of socio-technical inequality. These findings demonstrate that bias in ODR systems is not merely a technical issue but a multidimensional socio-legal phenomenon shaped by the interaction between algorithms, human cognition, and institutional contexts. The study contributes to the development of socio-legal and critical algorithm scholarship by emphasizing that justice in digital environments is process-oriented, relational, and context-dependent. Practically, the findings highlight the need for more transparent, explainable, and inclusive AI systems, alongside the preservation of meaningful human oversight. From a policy perspective, regulatory frameworks should address not only technological performance but also fairness, accountability, and accessibility. Future research is recommended to adopt comparative and longitudinal approaches to further examine evolving perceptions of justice in AI-driven dispute resolution systems.