The moral inversion of administrative AI: A critical interpretation on regime values

Abstract

Administrative AI is leading governance towards autonomous problem-solving without human intervention, which substantially challenges the traditional approach that public officials use to legitimize their actions. How can we ascertain the responsibilities and accountability of public officials if they are not involved in the autonomous AI administration? This question remains unaddressed in the traditional literature on administrative ethics, especially in the context of moral inversion. This article argues that regime values, which refer to the fundamental polity principles, can guide public officials in understanding and navigating moral inversion. This article first articulates the specific legitimacy challenges created by administrative AI, considering both traditional and new moral inversions. Then, based on the ethics triangle model, it interprets the specific dilemma of regime values through the dialectic relationship among consequences, virtues, and principles. Finally, this article suggests expanding traditional constitutional approaches by introducing ethical anchors regarding the moral inversion. With these anchors, the interaction between AI and public officials will have a new foundation to reach a new type of checks and balances. This study contributes to the traditional regime values theory and makes future analysis of regime values more consistent in the digital age.