Artificial Intelligence and judicial decision-making

revisiting the Kelsen-Klug debate on the applicability of logic to law

Abstract

The Kelsen-Klug debate (1959-1965) regarding the applicability of logic to law culminates in the proposition of whether judicial decisions could be made by computers. Klug sought to assimilate what he called normative calculus to mathematical calculus, arguing for the possibility of a logical inference operation from a general norm that would result in the logically necessary content of an individual norm – an operation that computers could perform. Kelsen, however, believed that logic does not apply to legal norms and emphasized the central role of the act of will in the creation of law. This provides a framework that allows for the present-day conclusion that artificial intelligence cannot replace judges. Therefore, it can be concluded that while artificial intelligence may eventually be able to replace human intelligence in the preparation of judicial decisions, it cannot do the same with regard to the “policy of law problem” that is realized in the act of will of the holder of legal power.