Algorithmic Decision-Making and the Ambiguity of Transparency in EU Digital Regulation

Abstract

This paper explores the concept of algorithmic transparency, which aligns closely with algorithmic explainability. After highlighting the challenges of translating technological rules from natural to machine language and presenting attempts to define some measurable indicators of transparency, it addresses a few inconsistencies in the recent EU regulations on data protection, digital services and AI systems. Furthermore, it examines notable rulings from the highest Italian courts concerning the use of algorithms in public administration decision-making, pointing out existing uncertainties. It concludes that it would be necessary to produce measurable standards at a European level to be applied to the use of artificial intelligence systems by public and private entities, to effectively evaluate their transparency and fairness.