LIABILITY FOR DAMAGES CAUSED BY ARTIFICIAL INTELLIGENCE—A NEW SPECIAL CASE OF LIABILITY IN WHICH TRADITIONAL RULES PROVE INSUFFICIENT AND SOMETIMES INAPPLICABLE
LIABILITY FOR DAMAGES CAUSED BY ARTIFICIAL INTELLIGENCE—A NEW SPECIAL CASE OF LIABILITY IN WHICH TRADITIONAL RULES PROVE INSUFFICIENT AND SOMETIMES INAPPLICABLE

Summary/Abstract: The adoption at the European Union level of a regulatory framework for artificial intelligence, robotics, and related technologies, based on ethical principles and legal obligations, represents a new challenge for a paradigm shift in the traditional institution of civil liability. Over time, economic and social transformations have led to new interpretations of this liability, in an effort to facilitate redress for injustices caused in society and to better protect victims. Thus, we have witnessed the decline of subjective liability—which is contingent upon proving the perpetrator’s fault—and the emergence of a new perspective toward strict liability, independent of any fault, based on an analysis of the abnormality of harmful conduct. The issue of compensable damages has undergone extensive development and diversification, particularly regarding bodily injury and moral damages. Moreover, loss of opportunity has been recognized as “compensable” to the extent that the likelihood of its occurrence is real and serious. However, establishing liability for damages caused by AI proves to be far more complex, emerging as a new special case of liability in which traditional rules prove to be insufficient and, at times, inapplicable. This study aims to analyze the specific aspects of this potential civil liability, starting with a review of recent European Union regulations applicable to AI, in order to address the difficult issue of identifying the liable party as AI technology becomes increasingly autonomous.

• Details
• Contents