CIVIL LIABILITY FOR DAMAGE CAUSED BY ARTIFICIAL INTELLIGENCE: THE MODERN EUROPEAN APPROACH
Yurii Burylo · Entrepreneurship Economy and Law · 2022
The purpose of the article is to outline the main theoretical concepts and current legislative initiatives concerning tort liability for damage caused by artificial intelligence (AI) in the European Union. Research methods.The methodology of this article relies on such research methods as analysis and synthesis, as well as comparative method. Results.Artificial intelligence poses a challenge to existing tort law, as it can cause damage acting independently and, at the same time, it is not regarded as a legal entity like natural and legal persons.In theory, tort liability for AI-related damage may be viewed as vicarious liability, strict liability, or fault-based liability.There is also a theoretical possibility of granting legal personality to autonomous AI systems, thus making it possible to hold them directly liable for the damage they cause.However, this approach does not have much support at the moment, even though it cannot be ruled out in the future.Considering the legislative initiatives of the European Parliament, the most probable approach to civil liability for AI-related damage in the EU will be based on the assessment of risk posed by different AI systems and will include strict liability of the operators of high-risk AI systems as well as fault-based liability of the operators of other AI systems which are not classified as high-risk. Conclusions.At the theoretical level, it is possible to approach the issue of civil liability for AI-related damage using the concepts of vicarious liability, strict liability, including product liability as well as faultbased liability.At the practical level, it is most likely that the European approach to developing legislation on civil liability for AI-related damage will be based on the assessment of risk and therefore will include a combination of strict liability for damage caused by high-risk AI systems and fault-based liability for damage caused by other AI systems that are not regarded as high-risk.In the near future, the possibility of granting legal personality to autonomous AI systems with for making them liable for damage does not seem realistic, although it cannot be ruled out in the long run.