Measuring Judicial Accountability in the Algorithmic Era: Juridical Implications of Using Judicial Assistants and Black Box Risks in Constructing Judgment Arguments

Abstract

Although the Supreme Court of the Republic of Indonesia has adopted Artificial Intelligence technology for administrative functions, the potential use of Artificial Intelligence as a Judicial Assistant in drafting judgment arguments triggers serious dogmatic concerns. These concerns relate to the degradation of human legal reasoning amidst a national legal vacuum (rechtsvacuüm). This study aims to deconstruct the concept of judicial accountability, which fails to address algorithmic error. Furthermore, this study tests the validity of Black Box-based rulings vis-à-vis the principle of reasoned decision in Law Number 8 of 1981, and formulates a preventive regulatory model. Utilizing a normative-juridical research method and a comparative law approach regarding regulatory frameworks in the European Union, the United States, and China, this study finds that conventional legal doctrines face a liability gap due to the unforeseeable autonomous behavior of Artificial Intelligence. The analysis indicates that reliance on algorithms with opaque characteristics—as demonstrated by the COMPAS case in the United States—fundamentally violates the defendant’s right to explanation. This potentially triggers “the death of standards.” In this condition, judicial discretion is replaced by the rigidity of machine micro-directives. Furthermore, the practice of relinquishment by judges for the sake of administrative efficiency threatens independence and judicial wisdom. This study concludes the urgency of adopting a hybrid regulatory model integrating technical efficiency with the strict User Control principle from the European Ethical Charter. This serves to ensure technology remains a human-supervised servant of justice, not a master dictating rulings.