Abstract

Jurisprudence has assumed, until recently, the idealized individual rationality of apex judges. Nevertheless, they do not solve cases alone, necessarily engage in oral deliberation, or are ideally wise. Instead, institutional procedures, caseloads, and epistemic constraints influence how courts solve cases. However, it is unclear what epistemic standards could replace this idealized model. This article proposes a framework that addresses this issue based on a virtue reliabilist approach and offers a solution to the problem concerning a proper understanding of the epistemic agency of apex courts. We propose a hybrid approach combining limited artificial intelligence automation based on precedent on routinary issues with a reflective approach to controversial ones. In this framework, automatized tools and judges are proxies that filter routinary from exceptional issues, paving the way for voting and realistic deliberation that is epistemically adequate