The Digital Identity of the Perpetrator and Achieving the Purpose of Punishment

The identification of the perpetrator, as the subject of a criminal offense, represents, in addition to the place, time, victim and specific criminal offense, the basis for initiating criminal proceedings, whose ultimate goal is to determine the guilt of the perpetrator, imposing a penalty or other criminal sanction and achieving the purpose of punishment. A person as a perpetrator and his real identity make him the subject of a criminal offense, committed in a real (or virtual) space and time. The purpose of punishment is defined in order to be achieved in relation to the perpetrator and other potential perpetrators. However, in the postmodern era in which we live, the question arises of whether the purpose of punishment can be achieved in relation to digital identities or autonomous systems of artificial intelligence (AI) as perpetrators of criminal offences in the virtual space. Can then the purpose of punishment be achieved by punishing a natural person with a real identity in relation to one or more digital identities or characters in the virtual space? A special problem related to criminal liability of a digital identity arises with AI systems that take autonomous actions in both the virtual and real space. In the paper, the author raises the issues of the criminal liability of autonomous AI systems in the context of the responsibility of legal entities (similar to the criminal liability of legal persons), types of possible penalties for AI systems and the need to determine a special, new purpose for sentencing such entities.