Civil liability in AI diagnostic analysis: theories and regulatory challenges

Abstract

The aim of this article is to raise thoughts about possible adaptations to the classic theories of civil liability applied to health, in the face of the use of medical devices with Artificial Intelligence (AI). The article seeks to situate readers in relevant aspects of the topic and to help with a general understanding of the proposed issue. The results of the study indicate the problem of the emerging behavior of machines, as well as the lack of definition of what would be the diligence expected of a doctor in the context of diagnostic analysis. It is argued that it is unreasonable to compare the data processing capacity of an artificial intelligence with the ability of a human doctor in the investigation of guilt. The final considerations reveal some changes in the interpretation of the assumptions of civil liability.