Artificial intelligence, Algorithmic Power and the Transformation of Democracy

Abstract

Artificial intelligence (AI) is reshaping the infrastructure of democratic life, simultaneously extending the reach of surveillance and reorganizing the spaces in which citizens deliberate. This paper offers a theoretical synthesis of two traditions rarely placed in sustained dialogue with one another in the AI-and-democracy literature: Michel Foucault's theory of disciplinary power, biopolitics, and governmentality, and Jürgen Habermas's theory of the public sphere and communicative action. Drawing on a systematic review of peer reviewed and scholarly sources, from an initial pool of 100 identified works twenty were retained for close synthesis, ten engaging Foucauldian surveillance theory and ten engaging Habermasian theories of political participation and deliberation. Six cross-cutting themes emerged: the transformation of the panopticon into algorithmic governmentality; the biopolitical management of populations through predictive data; the entanglement of surveillance capitalism with the colonization of the lifeworld; the algorithmic restructuring of the public sphere; the promise and limits of AI as a deliberative mediator; and the prospects for democratic resistance and accountability. The synthesis suggests that AI does not simply strengthen or simply threaten democracy; it does both at once, disciplining citizens through opaque data infrastructures while also creating new, contested spaces for rational-critical discourse. The paper concludes with implications for democratic theory, technology governance, and future research.