Two Years after the Adoption of the EU AI Act (Regulation (EU) 2024/1689, 13 June 2024): Impacts on Legal Practice through Selected Italian Judgments—Stricter Diligence Obligations and the Concept of Pedagogical Damages

Abstract

Two years after the adoption of Regulation (EU) 2024/1689 of 13 June 2024 (“AI Act”) and shortly after the entry into force of Italian Law no. 132 of 23 September 2025, the first body of Italian case law on the so-called “algorithmic hallucinations”—i.e. fictitious judicial precedents generated by Large Language Models and incorporated, without verification, into pleadings—makes it possible to assess the impact of the new normative framework on the legal profession. This paper systematically reconstructs six rulings issued between March 2025 and March 2026 (Florence, Latina, the Lombardy Administrative Tribunal, Ferrara, Syracuse and Mantova) and reads them against three interpretative coordinates: i) Article 14 of the AI Act, which, mutatis mutandis, lays down the principle of effective human oversight over high-risk AI systems and the duty to counter “automation bias”; ii) Article 13 of Italian Law no. 132/2025, which confines the use of AI by intellectual professionals to merely instrumental purposes and preserves the prevalence of human intellectual work; iii) the theoretical framework of punitive—or, more properly, “ pedagogical ”—damages, whose plurifunctional rationale (afflictive, educational, deterrent, compensatory and symbolic) sheds new light on the application of Article 96, paragraphs 3 and 4, of the Italian Code of Civil Procedure as consistently activated by the surveyed case law. The conclusion is that the Italian judiciary, anticipating the gradual application of the AI Act, has already developed a coherent doctrine of professional diligence which is best understood through the lens of pedagogical, rather than merely punitive, sanctioning.