The Theory of al-Ḍamān al-Dhakī in Civil Liability for Damages Caused by Artificial Intelligence Systems: An Applied Fiqh Study

Abstract

This study examines civil liability for damage caused by autonomous and semi-autonomous intelligent systems in light of the difficulties posed by algorithmic complexity and model opacity in proving wrongful conduct and establishing causation between human contributions and resulting harm. It aims to develop a contemporary jurisprudential framework that responds to the distinctive characteristics of intelligent technologies while preserving the general foundations of liability in Islamic jurisprudence and avoiding the conversion of technological complexity into a basis for strict liability. The study adopts an inductive and analytical methodology, supported by a maqāṣid-based approach and relevant judicial applications. It develops the theory of “Smart Liability” on three complementary foundations: wrongful causation (al-taʿaddī bi-al-tasabbub) as the substantive basis of liability; liability based on presumption (ḍamān al-tuhmah) as an exceptional evidentiary mechanism where direct proof is impeded by technical opacity; and innovative joint liability (al-taḍāmuniyyah al-ibtikāriyyah) for cases involving multiple human contributions whose causal effects cannot initially be disentangled, while preserving each party’s right to disprove causal involvement. The study further identifies defenses, including the victim’s conduct, absence of causation, and external causes or force majeure, with due regard to the technology’s nature and disclosed limitations of use. The theory is examined through applications involving advanced driver-assistance systems, medical AI, facial recognition, generative AI, deepfakes, and AI-enabled cyberattacks. The findings indicate that Smart Liability can address evidentiary gaps while protecting victims without imposing liability merely on the basis of developing or operating intelligent systems.