Cybercrimes Committed Using Artificial Intelligence in Algerian Legislation: A Critical Analytical Study of Substantive and Procedural Challenges and Prospects for Legislative Reform

Resumen

This study addresses a delicate legal problematic emerging from the accelerated evolution of artificial intelligence (AI) technologies: the capacity of the Algerian legislative framework to accommodate cybercrimes committed through autonomous, self-learning systems. The convergence of deep learning algorithms with cyberspace has generated a novel criminal typology characterized by functional autonomy, high velocity, and transnational reach. These features collide with the classical structure of criminal theory, grounded in the duality of actus reus and mens rea, as well as the principle of legality of offenses and penalties. Employing a critical-analytical methodology applied to Algerian legislative texts—notably the Penal Code, Law No. 09-04 on the prevention and combating of ICT-related crimes, and Law No. 18-07 on personal data protection—supplemented by comparative references to the European framework (particularly the 2024 EU Artificial Intelligence Act) and French legislation, the study concludes that the current Algerian regime suffers from a dual conceptual and procedural deficit. The substantive deficit is linked to the rigidity of legislative terminology in the face of synthetic generation phenomena (e.g., deepfakes), while the procedural deficit is connected to the inability of digital evidence mechanisms to address the algorithmic "black box" problem. The study proposes structural legislative reforms, including the introduction of a specific aggravating circumstance linked to AI use and the expansion of indirect criminal liability in accordance with the "Safety-by-Design" principle.