The Question of Reforming the Slovenian Criminal Law System due to the Development of Artificial Intelligence

Abstract

In this paper, we first analyzed the existing criminal law system, which does not impose criminal liability on Artificial Intelligence, in relation to modern theories that allow for this possibility, and tried to determine whether it is reasonable to introduce criminal responsibility of Artificial Intelligence. Furthermore, we have tried to establish whether the existing catalog of criminal offenses requires an adaptation or supplementation related to usage of Artificial Intelligence as a tool for perpetrating criminal offenses. Artificial Intelligence systems are already carrying out activities that in the past were purely human, and legal responsibility for harmful acts of such a system is becoming a key issue. In the literature, there are some concerns about whether to grant these artificial entities the legal subjectivity and hence the possibility of attributing responsibility for crimes (Hallevy, 2010, 2015; Music, 2016; Solum, 1992). This paper presents arguments which oppose this option. Artificial Intelligence can be programmed to function in accordance with the rules, but it does not understand the deeper meaning of these, and at the same time, is not capable of hermeneutical judgment (Eidenmiiller, 2016). From an ontological point of view, it can be concluded that the law, and rights and obligations associated with it is a human construct, and reflect what people perceive as a prerequisite for the normal functioning of society. At the same time, the law reflects what we believe is the essence of humanity and the essence of what it means to be human. The granting of rights and obligations to Artificial Intelligence would dehumanize the concepts that are originally human and represent the core of the human essence (Eidenmiiller, 2016). However, Artificial Intelligence at the current stage of development does not possess awareness and will. Therefore, the cause of the damage that arises from the operation of Artificial Intelligence must be found in the person who stands behind it. This can either be a manufacturer, a programmer, or an end-user. We believe that a person acts as a perpetrator when he "abuses" Artificial Intelligence by committing a criminal act through it. In this case, Artificial Intelligence is merely his tool. The range of criminal offenses that can be committed through Artificial Intelligence is broad. In the paper, we found that there is no need for new incriminations since the behaviour of a person abusing Artificial Intelligence can be subsumed under existing criminal offenses.