Algorithmic Governance and Legal Protection of Citizens in AIBased Administrative Decisions

Abstract

This study examines the vagueness of norms in the legal protection of citizens against AI-based administrative decisions within the Indonesian administrative law system. The increasing use of Artificial Intelligence in public administration, particularly in decision-making processes such as social assistance distribution, licensing, and digital public services, raises significant juridical challenges related to legality, transparency, accountability, and access to justice. The central legal issue concerns whether algorithmic outputs can be classified as State Administrative Decisions, the absence of clear standards for explainability and accountability in algorithmic processes, and the lack of effective legal remedies for citizens affected by automated decisions. This research employs a normative juridical method using statute, conceptual, and case approaches, analyzing relevant legal frameworks including the 1945 Constitution, the Government Administration Law, the Public Service Law, the Personal Data Protection Law, and the Public Information Disclosure Law. The findings indicate that Indonesian positive law has not adequately accommodated the characteristics of AI-based administrative decision-making, resulting in legal uncertainty and potential violations of fundamental legal principles such as due process of law and equality before the law. This study proposes a prescriptive legal framework by recognizing AI-based decisions as reviewable administrative acts, strengthening transparency through explainable and auditable AI systems, and reformulating procedural mechanisms to ensure effective legal remedies and protection of citizens’ rights.