Abstract

This article analyses the applicability of the indirect perpetration model as a basis for personal attribution to natural persons who employ AI-powered devices' technical capabilities, and it adopts a favourable stance. In this analysis, the article introduces, as an initial conceptual approach, the category of actants (comprising such devices). This challenges the modern criminal law's bipartite conceptual framework, which solely distinguishes between subjects and objects. It is proposed that this category of actants may serve as an effective instrument for reinterpreting criminal law structures when addressing the issue of criminal liability arising from the involvement of intelligent machines.