Özet:

In this article authors discuss existing ideas about liability of artificial intelligence based on guilty and strict approaches to defining the elements of civil liability in the Russian Federation and European Union. These approaches have drawbacks, which are, first of all, in the excessive limitation of the development of innovations, and with low efficiency in achieving the goals of civil legal responsibility and the implementation of its functions. The risk-based approach proposed by the author to the determination of the elements of civil liability for the actions of artificial intelligence is intended to neutralize the named drawbacks. Based on the analysis of the spheres of application and artificial intelligence technology, the risk-based approach allows a more efficient and flexible approach to the definition of the subject of responsibility, its types and limits, ensuring a balance between the development of innovation and the goals of civil liability. As a result of the study, the author’s definition of a risk-based approach to civil liability for the actions of artificial intelligence has been given, its features, elements have been disclosed, and its advantages over existing approaches to civil liability have been demonstrated.

Anahtar Kelimeler: