Criminal Liability of Artificial Intelligence in the Iranian Legal System: A Legal–Jurisprudential Analysis of Emerging Challenges and a Comparison with Selected Legal Systems

Abstract

The expansion of artificial intelligence technologies, particularly learning, autonomous, and decision-support systems, has confronted criminal law with fundamental questions. The principal issue is determining which person or persons should bear criminal liability when the operation of an intelligent system results in the commission of an offence, bodily, financial, or reputational harm, a violation of privacy, or the generation of criminal content. In Iranian criminal law, criminal liability is founded on principles such as the legality of crimes and punishments, the personal nature of criminal responsibility, legal capacity, volition, the mental element, imputability, and causation. In Imamiyyah jurisprudence, doctrines and rules such as direct perpetration, causation, destruction of property, liability for damage, the rule of no harm, the rule of inducement or deception, the prohibition against harming others, and the requirement to establish intent and free choice in criminal liability and civil liability also play a significant role in analysing harmful conduct. Employing a descriptive–analytical method and adopting legal, jurisprudential, and comparative approaches, this study examines the status of artificial intelligence within the Iranian legal system, analyses the challenges associated with its criminal liability in light of jurisprudential foundations and criminal-law principles, and briefly considers the approaches adopted by the European Union, the United States, and the United Kingdom. The findings indicate that, under the current legal framework, artificial intelligence cannot be recognised as possessing independent criminal liability in the conventional legal sense, because such systems lack rational agency, moral will, criminal intent, legal capacity, and the ability to undergo punishment. Nevertheless, the natural or legal persons involved in the design, production, training, marketing, supervision, and operation of an artificial intelligence system may, depending on the circumstances, be held liable on the basis of criteria such as control, foreseeability, fault, causation, and attribution. Accordingly, the enactment of comprehensive artificial intelligence legislation in Iran, with due regard to the principles of criminal law, jurisprudential rules, and the regulatory requirements applicable to high-risk technologies, is an undeniable necessity.