FROM PLATO’S CAVE TO THE MAZE OF MIRROS

THE CITIZEN-ROBOT AND THE ETHICAL USE OF ARTIFICIAL INTELLIGENCE IN ENVIRONMENTAL LEGAL AUTOMATION

Abstract

This article analyzes artificial intelligence through a genealogical lens that reveals a field marked by epistemological caves—where shadows of neutrality conceal corporate rationalities—and by institutional labyrinths in which the appearance of innovation coexists with technical opacity and power disputes. In this context, the metaphor of Ariadne’s thread is proposed as a critical guide capable of integrating historical, philosophical, and legal dimensions to understand the impacts of AI on the Brazilian justice system. The analysis examines the national regulatory framework, highlighting the LGPD, Bill No. 2,338/2023, and CNJ Resolution No. 615/2025, which introduce principles of transparency, human oversight, explainability, and risk management as structural conditions for institutional uses of AI. The study also investigates the experimental phase of AI in the Judiciary, marked by technological fragmentation and risks of digital coloniality. Within this horizon, it develops the architecture of the “Citizen-Robot,” conceived as a public technology guided by the Constitution and designed to provide qualified support in environmental protection, based on RAG, open-source models, and auditable governance. The article concludes that the genealogy of AI provides the critical tools necessary to develop responsive, ethical, and teleologically aligned solutions that strengthen citizenship, democratic justice, and the effectiveness of environmental rights.