Due process by design: Enhancing fairness and trust in AI decision-making
Year: 2025
Type: article
Abstract: This article explores the intersection of socio-normative procedural justice theory and the growing use of AI in decision-making. While AI systems offer substantial advantages, they also raise significant concerns related to discrimination, fairness, and public trust, fueling global demands for “trustworthy,” “responsible,” and “accountable” AI. However, translating these broad ideals into concrete frameworks remains a major challenge. The article argues that procedural justice theory, which emphasizes the fairness of decision-making processes as a foundation for building trust in various social institutions, offers a compelling basis for addressing these concerns. While foundational procedural principles such as transparency have been extensively addressed in the AI context, a key component of procedural justice—the right to be heard, as part of a due process—is largely overlooked. Hearings allow individuals to actively participate in the decision-making process, enhancing personalization, accuracy, perceptions of fairness, and ultimately, trust. Building on value-sensitive design and human-centered technology approaches, the article introduces the concept of “due process by design.” It advocates embedding procedural elements, notably automated hearings, into the architecture of AI decision-making systems. Drawing on insights from socio-legal legacies and public policy, it proposes adaptive, real-time input mechanisms that emulate hearings as understood in traditional due process frameworks. These mechanisms can improve the accuracy and individualization of decisions while enhancing users’ perceptions of fairness and trust in AI decision-making processes. This interdisciplinary approach, linking socio-normative theory, public policy, and technology design, aims to strengthen fairness and public trust in civil society decision-making systems.
Author: Orit Fischman-Afori
Institution: College of Management Academic Studies
Language: English
Subfield: Safety Research
Field: Social Sciences
Domain: Social Sciences
Open Access status: gold
Estimated APC paid: $1,180
OpenAlex ID: W7109671987