Digital Person as Legal Fiction Under the AI Act, GDPR, and eIDAS

This article examines how the EU’s emerging regulatory triangle, Regulation (EU) 2024/1689 (the AI Act), Regulation (EU) 2016/679 (the GDPR), and Regulation (EU) 2024/1183 (eIDAS 2.0/EUDI Wallet), reconstructs the legal identity of individuals in algorithmic environments. It argues that, taken together, these instruments implicitly generate a functional legal fiction of the digital person: the composite techno-legal representation through which the natural person’s rights, attributes, and vulnerabilities are mediated by credentials, data traces, profiles, and AI-assisted decisions. Drawing on Roman law’s stratified concepts of persona, status, and caput and on modern theories of legal fiction, the article shows why treating AI as an autonomous legal subject would displace responsibility and fragment remedies, while treating individuals only as data subjects is normatively too thin for high-stakes digital governance. The paper proposes a twofold move, namely the normative personification of the individual in digital contexts and the use of transparent liability vehicles for certain AI systems, such as a limited electronic legal person (PJE), without ontological personality.