AI-DIGITAL HUMAN AS A NEW CLASS OF DIGITAL LAW SUBJECT: TYPOLOGY, CRITERIA OF LEGAL PERSONHOOD, AND A REGISTRATION MODEL IN DIGITAL ECOSYSTEMS
Abstract
This article argues for a new class of digital legal subject within AI legal-personhood doctrine — the ‘AI-digital human’ — largely unaddressed in domestic and international legal scholarship. Building on the author’s earlier three-class model of AI legal subjects and the institution of the AI-representative, the article distinguishes two different modules of the AI-digital human: Module 1 (Digiperson), a digital human as a functional reflection of a specific natural person, built on ‘red-group’ identification attributes and digitised access credentials; and Module 2 (autonomous Digihuman), reflecting no real natural person, its attributes formed referentially and tied to a functional role within a digital ecosystem. The article shows that equating a digital human with a natural person in the full scope of rights and obligations is a doctrinal error: legal status should follow an age-based logic (child—adolescent—adult) rather than mirroring the prototype’s status. Drawing on a comparative analysis of Regulation (EU) 2024/1689 (AI Act) and Regulation (EU) 2024/1183 (eIDAS 2.0), the article substantiates the need for a Digital Humans Registry as an institutional legitimation mechanism for both modules, and formulates draft articles for the Ukrainian Law on Artificial Intelligence.
