Legitimacy Debt, Neuro-Sovereignty, and Friction Rights: Reconstructing the Social Contract for AI-Mediated Governance

Public institutions are adopting AI decision systems at the same time that public trust in those institutions is eroding, and the standard remedy-greater transparency-no longer suffices. This paper develops a theory of legitimacy for AI-mediated governance and proposes institutional mechanisms to defend it. It formalizes legitimacy debt as the accumulated, unrepaired violation of procedural justice-fair process, voice, and equal respect-in AI-mediated systems, and distinguishes three layers: economic, symbolic, and neurological. It then introduces Neuro-Sovereignty, a governance-level right to mental self-authorship in the face of predictive, affective, and attention-shaping systems, and argues that it is distinct from privacy, data protection, freedom of thought, and the individual neurorights now entering constitutional law. To operationalize procedural justice inside automated environments, the paper develops Friction Rights-a guarantee of slow, contestable moments in high-stakes automated flows-and Verification Pauses, specifying who may invoke them, what must be logged, what explanation is owed, and how override works. It connects these to emerging oversight frameworks for agentic AI through a four-pillar governance pattern and positions Friction Rights as a minimum legitimacy condition. The paper closes with a normative and empirical research agenda, including candidate measures for legitimacy debt and Neuro-Sovereignty and testable hypotheses linking genuine friction to perceived legitimacy.