Liability for damages caused by artificial intelligence in medicine

Abstract

The article is devoted to two problems. First of them is the catalogue of entities liable for harms caused of the functioning of artificial intelligence in medicine. Second of them is the principles on which this liability is based. The study presents comments on the biding law, which show that the software, regardless of its advancement level, cannot be subject of rights and obligations under civil law - including compensation ones. The text considers the need to expand the group of entities with legal capacity to include intelligent software. This analysis led to the conclusion that such a change in the legal status would not be advantageous. The following part of the study presents the role of a medic who uses artificial intelligence in performing professional activities and the role of a patient in making a decision to provide a health service with the use of the described digital tools - in the context of the premises for liability for damages. The considerations contained in the last of essential parts of the article refer to the liability for harms of the software producer, as well as other persons who have an influence on the content of the software code. The study ends with a summary, in which the postulate de lege ferenda was submitted, that the liability of a medical professional using artificial intelligence in providing health services should be based on the principle of presumed fault, and the liability of the software producer or liability of other entities interfering in software code should become a strict liability by extending the definition of the product under Polish Civil Code.