Thinkers whose frameworks illuminate this section.
Habermas's communicative action theory and public sphere analysis are the most direct theoretical complement to Rosanvallon — both trace how democratic legitimacy requires not just voting but ongoing deliberative exchange, and Habermas's concept of the colonization of the lifeworld maps directly onto AI's transformation of public discourse.
Arendt's distinction between labor, work, and action — and her insistence that political action in the public realm is irreducible to technical mastery — provides the philosophical foundation for Rosanvallon's demand that AI governance belong to the political, not the technical, domain.
Jasanoff's co-production framework — showing how scientific and technological knowledge is built simultaneously with social order — directly supports Rosanvallon's argument that AI governance cannot be separated from democratic legitimacy; who counts as an expert is itself a political decision.
Mazzucato's analysis of how public investment drives innovation — and her argument that mission-oriented institutions can direct technological development toward democratic goals — provides the institutional design complement to Rosanvallon's framework of what democratic governance of AI could look like.
Unger's democratic experimentalism — his insistence that democratic societies can remake their institutional arrangements and are not bound by existing structures — is the practical extension of Rosanvallon's reflexive democracy; both argue that institutional invention is the fundamental democratic capacity.
Polanyi's account of the double movement — market expansion generating protective counter-movements demanding institutional embedding — predicts exactly the democratic demand for AI governance that Rosanvallon's framework theorizes; the AI transition is the latest iteration of the market-society dialectic Polanyi traced.
Winner's argument that artifacts have politics — that technical systems embed values and power relations that require democratic scrutiny — is the philosophy of technology complement to Rosanvallon's political science; together they ground the claim that AI governance is inescapably political.
Stiglitz's analysis of information asymmetries — and his argument that markets without democratic correction produce inequality and dysfunction — provides the economic framework for why AI's opacity is not just a technical problem but a democratic emergency requiring institutional response.
Scott's concept of legibility — how states and platforms simplify complex realities to make them governable, at the cost of local knowledge — maps directly onto AI systems' reduction of human behavior to training data; his metis (practical knowledge) is what Rosanvallon's proximity legitimacy protects.
Weber's typology of legitimate authority — traditional, charismatic, rational-legal — and his analysis of bureaucratic rationalization provide the sociological foundation for Rosanvallon's multi-dimensional legitimacy framework; AI governance represents a new form of rational-legal authority without democratic roots.
Berlin's distinction between negative liberty (freedom from interference) and positive liberty (capacity for self-governance) illuminates the tension in AI governance between technical safety constraints and democratic empowerment; Rosanvallon's counter-democracy is an argument for positive liberty at collective scale.
Luhmann's systems theory — particularly his analysis of how differentiated social systems develop autonomous logics that resist external interference — provides the theoretical framework for why AI companies operating as self-referential systems resist democratic accountability, and why Rosanvallon's institutional relays are necessary.
