The criminal liability of Artificial Intelligence entities – from science fiction to legal social control

Abstract

One of the most important issues related to artificial intelligence is whether thinking machines are subject to criminal law control like the other legal subjects? In this article an attempt is made at suggesting a legal solution to the problem of criminal liability of artificial intelligence. On the grounds of an analysis of its essence, three possible models are proposed: the perpetration by means, the natural and foreseeable consequences and the direct liability ones. In order for criminal pro section against artificial intelligence to be possible, these three models could be applied either separately or in the relevant combination.