Artificial intelligence and digital technologies in public administration: a constitutional and legal dimension

Abstract

This article presents a comprehensive constitutional-legal analysis of the use of artificial intelligence and other digital technologies in public administration. The study aims to define the constitutional and legal foundations for their application, identify key risks to human rights and freedoms, and substantiate a differentiated model of legal safeguards based on the intensity of digital influence on legally significant outcomes. The relevance of this issue stems from the gradual shift from using digital tools merely for technical support to employing them in processes involving data analysis, the formulation of recommendations, the determination of parameters for official decisions, and the automated adoption of legally significant decisions. The study employs formal-legal, comparative-legal, and system-structural methods, as well as the method of legal modeling. It analyzes constitutional principles and norms of national legislation, contemporary European and international approaches to the regulation of artificial intelligence, and the specific aspects of using automated systems in the activities of public administration. The article establishes that the constitutional and legal significance of digital technology is determined by the extent of its impact on the exercise of public authority and the legal status of the individual. It is argued that a digital system cannot serve as an independent source of public competence; rather, its software architecture and settings must comply with legally prescribed limits of authority and cannot effectively alter the substance of legal regulation. The article analyzes the requirements of legality, legal certainty, equality and non-discrimination, privacy protection, accountability, due process, and effective legal protection within the context of the algorithmization of public administration. The scientific novelty of the study lies in substantiating the principle that constitutional and legal safeguards should be strengthened in proportion to the intensity of digital influence. Building on this principle, the article proposes a four-level model of digital influence and develops a digital influence intensity test based on four criteria: the functional role of the system, the nature of the legal consequences, the characteristics of the data used, and the extent to which meaningful human review is genuinely available. The study further substantiates a set of safeguards applicable in cases of substantial digital influence, including the right to be informed about the significant use of artificial intelligence, a legally sufficient explanation of the decision, meaningful human review, a prior impact assessment concerning constitutional rights, digital auditing, and effective mechanisms for review and appeal. It is concluded that it is advisable to establish a differentiated model of legal regulation for the use of artificial intelligence in public administration in Ukraine, wherein the scope of legal safeguards is determined by the digital system's actual role in generating a legally significant outcome. This approach makes it possible to reconcile the efficiency of public sector digitalization with the requirements of legality, accountability, and adequate human rights protection, without imposing uniformly stringent regulatory requirements on technologies that have varying degrees of legal impact.