Rabia Sağlam
Kocaeli University Law School, Turkey
This paper argues that equitable legal reasoning is not an exclusive property of human judges. To investigate whether judges’ perceptions can be regarded as a fundamental and special prerequisite for equitable reasoning, I discuss the virtue jurisprudence thesis. Then, using the aporetic logic of deconstructive justice, I outline a framework for the prospective content of equitable reasoning for human judges and future artificial intelligence (AI) judges. However, despite the optimistic and visionary claims about AI judges’ ability to make decisions in a similar way to human judges, pervasive skepticism exists. This essay critically analyzes this skepticism by focusing on the possibility of programing algorithmic models to encode values associated with Aristotle’s concept of epikeia. To demonstrate how algorithmic reasoning can be harmonized with value-based principles, a case-based algorithmic model is presented as a representative example. The goal of this example is to challenge the assumption that human judges have an inherent advantage in (equitable) legal reasoning. To illustrate that the reasoning of human judges is not always equitable, I present an example of an inequitable decision by the Turkish Supreme Court of Appeals. These examples serve to justify my two claims. First, there is no compelling reason to favor the ‘virtuous’ and ‘legal’ reasoning of human judges over the ‘algorithmic’ reasoning of AI judges. Second, the legal responsibility to consider the particular nuances of each case—the central element of equitable reasoning—can be seen as a quality attributed to both human and AI judges.
コメント