The Purpose of Punishment in Crimes Involving Human-Robot Interaction

The increasing use of artificial intelligence in human activities raises new questions concerning the purposes of criminal punishment. This paper examines how human–robot interaction affects the traditional justifications of punishment by focusing on the role of AI in the cognitive, decisional, and operational phases of criminal conduct. These forms of technological mediation modify the relationship between the individual, the wrongful act, and the resulting harm, thereby influencing the assessment of responsibility and the meaning attributed to punishment. Against this background, the paper considers both the retributive and the consequentialist theories of punishment. It is argued that, from a retributive perspective, AI may weaken the correspondence between moral culpability and punishment by altering the connection between the offender's will, conduct, and harmful consequences. At the same time, from a consequentialist perspective, the preventive functions of punishment—including deterrence, rehabilitation, and incapacitation—may become less effective where the risk of future offending depends not exclusively on the individual but also on the interaction with AI systems. The paper does not suggest abandoning traditional theories of punishment. Rather, it argues that the transformations introduced by AI require a reconsideration of their application in order to preserve both the legitimacy and the effectiveness of criminal punishment in a technologically mediated environment.