Displacement, Dilemma, and Exit: The Ethical Stakes of Algorithmic Decision-Making

Abstract

Algorithmic governance reshapes the exercise of authority by distributing decisions across code rather than concentrating them in identifiable human agents. This paper argues that three functions traditionally served by personified authority come under structural pressure: the anchoring of accountability, the adjudication of value conflicts, and the reproduction of the conditions for collective self-governance. The ethical stakes concern the capacity of affected persons to hold decision-makers to account, to participate in forming the norms that govern them, and to act as members of a self-governing polity. To make these dynamics visible, the paper develops three diagnostic concepts through blockchain governance and applies them to AI-based decision-making. Accountability anchor displacement names the condition in which distributed architectures dissolve or functionally bypass the subject to whom the demand for an account must be addressed. The Althusian dilemma describes polycentric normative orders that lack a recognized mechanism for adjudicating value conflicts. The erosion of civic virtue through exit captures how governance architectures that make exit cheaper than voice systematically weaken the institutional conditions for collective self-governance. The paper draws on the theory of Althusius, Hirschman, and Machiavelli, alongside contemporary scholarship on the EU Artificial Intelligence Act and the Council of Europe Framework Convention on AI. The core claim is that algorithms cannot cancel the need for institutional anchors that hold accountability in place, adjudicate value disputes, and sustain civic participation.