Abstract

The article attempts to analyze the matter of civil law liability for actions performed with the use of artificial intelligence (AI), with regard given to the phenomenon of AI autonomy that should be considered symptomatic and legally challenging. The authors examine the matter in the light of liability regimes existing under Polish civil law, including strict and fault-based liability, analogies to liability for animals and references the product liability according to Council Directive 85/374/ EEC. The authors suggest that the existing regulations seem to be insufficient or only partially adequate. Making use of examples of the AI application in the field
of intelligent robots and autonomous vehicles, the authors emphasize the lack of comprehensive regulations that would determine principles of liability in instances where damage is linked to the use of new technologies and indicate difficulties in effective pursuing claims by injured persons. Liability for product seems to be a suitable starting point for mitigation of legal uncertainty, but due to concerns related to the involved detailed solutions, an amendment of the EU legal framework appears indispensable. The article also presents initiatives taken at the EU level to regulate issues related to the matter, including the outcomes of work of expert groups set up by the European Commission.