Governing the Perceptual Infrastructure

Public Administration in the Age of Algorithmic Perception

Abstract

Artificial intelligence systems increasingly shape not only what information citizens access but the interpretive frameworks through which they perceive public affairs, posing a governance challenge that public administration scholarship has yet to theorise. Through conceptual and normative analysis, the paper develops the construct of perceptual infrastructure – the cognitive and informational substrate of democratic deliberation – drawing on information theory, public administration and regulatory-governance scholarship, and tests it against the EU regulatory architecture. The analysis shows that the algorithmic construction of perceptual frameworks constitutes a distinct governance domain that the EU AI Act and the Digital Services Act do not reach, owing primarily to a regulatory omission rather than an implementation deficit: existing provisions take systems, use cases and identifiable harms as their object, not the cumulative, longitudinal construction of perception. An accountability framework is proposed – incorporating aggregate transparency, perceptual sovereignty as a citizen right, proportional responsibility and meta-perceptual literacy – with concrete implications for administrative accountability, regulatory capacity and democratic resilience.