The Civil Liability of Artificial Intelligence Applications: Between the Limitations of Traditional Liability and the Evolution of the Product Concept

Abstract

The expansion in the field of artificial intelligence (AI) has led to the emergence of complex legal challenges, particularly concerning civil liability for damages that may arise from these systems. This raises questions about the adequacy of traditional legal frameworks in addressing such technological transformations. This study examines civil liability for damages resulting from the use of AI applications by analyzing the suitability of contractual and tort liability rules, as well as the need to develop the legal concept of a product to enable the application of strict liability. The research employs a comparative analytical approach, focusing on three Arab countries: the United Arab Emirates, Iraq, and Egypt. Its aim is to conduct a comparative analysis that highlights legislative gaps and frameworks of civil liability through a review of relevant legislation and contemporary legal scholarship, particularly within the context of the European Union. The objective is to formulate a modern legal perspective that reflects the nature and challenges of artificial intelligence. The study clarifies that traditional civil liability rules were not designed to address damages caused by autonomous, dynamic technological systems like AI, where proving fault or causation presents a significant challenge. This is particularly due to the multiplicity of actors involved (developers, users, data providers, and manufacturers) and the overlap in their roles. The research further demonstrates that strict liability, especially under the framework of defective product liability, represents a viable alternative for compensating victims without requiring proof of fault, provided that a defect exists in the system or intelligent application.