Critical View of the Introduction of Artificial Intelligence in the Administrative Procedure: Reception of the EU AI Act in Italy

Abstract

This article offers a critical examination of Italy’s initial legislative efforts to transpose the European Union’s Artificial Intelligence Act. Employing a methodological approach of argumentative-critical commentary, the analysis dissects the bill’s text to uncover the Italian government’s underlying philosophy and regulatory strategy. The study argues that the proposed framework suffers from a significant lack of overall vision, mistakenly treating AI as a non-penetrating technological tool and reducing its risks primarily to data privacy concerns, all while prioritizing industrial growth. The article critiques key provisions, including the ambiguous “anthropocentric” principle, the superficial regulations governing AI in information, healthcare, intellectual professions, and the judiciary, and the creation of new criminal offences. It further analyzes preliminary projects for integrating AI into administrative justice and explores the profound procedural implications of extending AI into areas of technical discretion, highlighting the potential for a paradigm shift in judicial review. By juxtaposing the Italian approach with a documented case of algorithmic discrimination in the UK, the research underscores the concrete risks to fundamental rights and the rule of law. The conclusion emphatically calls for a more technically competent and critically aware involvement from legal scholars and practitioners to safeguard citizens from the uncritical and pervasive adoption of AI in public administration and justice, warning against the shortcomings of the current fragmented and constitutionally superficial regulatory proposal.