Criminal Liability in the Age of Autonomous Systems: Rethinking Mens Rea and Actus Reus

Abstract

Due to an increase in the ability of self-acting systems to independently perform their work in high-risk fields, such as in the sphere of transport, healthcare, national security, fundamental issues concerning criminal liability arise. Crimen denotes crime, and crime (to us), involves violation or breach of law which must be obeyed by state as well as citizen in the interest of the collective well-being and safety of the society as a whole, yet human conviction and deliberateness lie at the basis of the legalizing concepts of mens rea (guilt) and actus reus (the actual act). Nevertheless, AI systems are not conscious (according to the human notion of it), and are not guilty before people, and their deeds can be devastating. In this paper, the key elements related to the criminal law are researched in terms of the emergence of autonomous systems. It alludes to the fact that the manner of arrangement of the law should be altered to accommodate the newer forms of agency and presents the idea that how we structure our models of culpability will have to be reconsidered as machines grow in autonomy. In the absence of such a transition, criminal law will have either excessively punished the human actors or became insufficiently regulated over AI-caused harms.