Régimen de Invalidez del Acto Administrativo Discrecional Automatizado en Colombia y México: Opacidad Algorítmica y Crisis de la Motivación

Abstract

The increasing incorporation of artificial intelligence systems into public administration poses a structural challenge for contemporary administrative law, particularly when discretionary powers are automated. The replacement of human reasoning with inscrutable algorithmic models generates a phenomenon of technical opacity that nullifies the motivation of the act, threatening fundamental constitutional guarantees. Faced with this problem, the main objective of this research is to determine how the grounds for nullity and illegality should be redefined and applied in the legal systems of Colombia and Mexico when the administrative act has been automated. To this end, a qualitative methodology based on legal dogmatics and micro-comparative law is used, critically analyzing the regulatory framework and jurisprudence of both countries. The results show that classical judicial control tools are insufficient in the face of a lack of technological transparency. Consequently, the study proposes recognizing algorithmic opacity as an autonomous ground for substantial invalidity, accompanied by an urgent procedural adaptation and redistribution of the burden of proof to prevent the consolidation of new spaces of state power immunity against citizens.