Criminal liability for artificial intelligence crimes and the penalties stipulated for them in Iraqi and Algerian legislation.

Abstract

This research examines criminal liability arising from artificial intelligence crimes under Iraqi and Algerian legislation. It analyses the parties who may be held liable when harmful conduct occurs, particularly manufacturers, programmers, users, owners, and external actors, while also discussing whether liability may be attributed to the artificial intelligence system itself. The research further presents contemporary criminal-law approaches to the autonomy of intelligent systems, the limits of recognising them as legal persons, and the effects of such recognition on the elements of crime, causation, and criminal intent. Using an analytical and comparative methodology, the research identifies similarities and differences between Iraqi and Algerian legal frameworks and evaluates the adequacy of general criminal rules in addressing AI-related offences. It concludes that both systems continue to emphasise human liability, despite legislative shortcomings concerning multiple actors and autonomous algorithmic decisions. It recommends specific regulation, stronger technical oversight, specialised evidentiary rules, and appropriate sanctions. The issue extends beyond information security and the digital economy to include threats to the physical safety of individuals when it comes to autonomous physical systems, such as self-driving cars, surgical robots, or systems integrated into critical infrastructure like power, water, and telecommunications networks. In these cases, any malfunction or criminal exploitation could lead to direct loss of life, thus elevating the challenge to a matter of public security and human safety.