From Algorithmic Decision-Making to Administrative Legitimacy: Rethinking Legal Accountability in Digital Governance

Abstract

The growing use of algorithmic systems in public administration is transforming how public decisions are produced, justified, and contested. As automated and data-driven processes become embedded in administrative practice, traditional foundations of administrative authority, including human reasoning, procedural transparency, reason-giving, and institutional responsibility, are increasingly placed under strain. This shift creates new tensions between administrative efficiency and legal legitimacy, raising important questions about how accountability frameworks should adapt to opaque, distributed, and dynamic algorithmic decision-making processes. This paper examines the implications of algorithmic governance for administrative legitimacy and argues that existing legal accountability models are insufficient when decision-making authority is dispersed across public agencies, private technology providers, data infrastructures, and technical systems. Adopting a conceptual and normative analytical approach, the study identifies key challenges, including algorithmic opacity, fragmented responsibility, automation bias, limited contestability, and data-driven discrimination. In response, it develops a conceptual-normative accountability framework that links these challenges to corresponding legal risks and institutional responses. The framework emphasizes meaningful transparency, answerability, accessible contestation, effective human oversight, structured responsibility, auditability, and impact assessment. The analysis suggests that sustaining administrative legitimacy in the digital era requires more than regulatory adjustment. It requires lifecycle-based accountability mechanisms capable of reconnecting algorithmic decision-making with legality, procedural fairness, institutional responsibility, and public justification.