CONCEPT

Infocracy

Han's 2022 diagnosis of a governance regime in which political deliberation is progressively replaced by algorithmic management — democracy degenerating into infocracy, with optimization substituting for judgment.

Infocracy is Han's name for the contemporary political condition in which democratic deliberation — the messy, conflictual, genuinely plural process through which communities negotiate shared existence — is being replaced by algorithmic systems that present themselves as neutral optimization. The algorithm determines credit scores, parole decisions, insurance rates, university admissions, social media visibility. Each substitution is locally rational; the algorithm is more comprehensive, more consistent, less susceptible to the cognitive biases that distort human judgment. Aggregated across institutions and decades, the substitutions produce what Han calls infocracy: a regime in which the political question — what kind of society do we want to live in? — has been converted into a series of technical problems, each one solved by optimization, none of them addressed by deliberation. The infocratic order does not announce itself as a regime. It announces itself as efficiency, as evidence-based decision-making, as the elimination of the messiness of human judgment. And in doing so, it eliminates the political itself — the space where human beings, in all their plurality and imperfection, decide together what matters.

In the AI Story

Hedcut illustration for Infocracy — Infocracy

Han's argument draws on Hannah Arendt's distinction between the political and the technical. The political, in Arendt's sense, is the space of genuine plurality — where different people, with different values and interests, come together to negotiate the terms of their shared existence. Politics is not management. It is not optimization. It is the difficult, often inefficient process of arriving at decisions that reflect the genuine plurality of a community, which means arriving at decisions that are, by definition, suboptimal from any single perspective.

Algorithms cannot do this. An algorithm optimizes — finds the solution that maximizes a defined objective function. But the definition of the objective function is itself a political act, a choice about what matters, whose interests are weighted, and how. The algorithmic framing conceals this choice, presenting as neutral mathematics what is in fact a set of embedded values. The infocratic move is not to decide politically and then implement technically; it is to present the political decision as already-technical, already-determined by the data.

Han quotes Alex Pentland's dream of a divine eye — a global vision that would achieve true understanding of how society works. Han reads this not as scientific aspiration but as theological hubris: the fantasy of a god's-eye view that dissolves the plurality of human perspectives into a single algorithmic perspective. The dream is the fantasy that psychopolitical power requires: the comprehensive legibility of the population that makes optimization possible.

The 2026 AI governance debate is an infocratic event in Han's sense. The question of how to govern AI is being framed as a technical problem requiring expert solutions, with technology executives and algorithmic systems presented as the natural authorities. The political question — whose interests are being served, who bears the costs, what kind of society is being built — is systematically displaced by questions about optimization, safety protocols, and risk thresholds. The displacement is not neutral. It reproduces, at the scale of governance, the achievement society's preference for the technical over the deliberative.

Origin

Infokratie: Digitalisierung und die Krise der Demokratie was published in German in 2021, with the English translation appearing in 2022. The book appeared as Han's most direct engagement with the political dimension of the digital transformation, extending his earlier analyses of psychopolitics into the question of democratic governance.

The term infocracy is Han's coinage, combining information and kratos (rule, power). It names a form of power that operates not through coercion or persuasion but through the framing of decisions as technical rather than political — a framing so thorough that the political dimension becomes literally invisible to the subjects making the decisions.

Key Ideas

Politics replaced by systems management. Decisions taken on the basis of big data and AI present themselves as neutral when they are nothing of the kind.

Optimization is not deliberation. The algorithm finds solutions; it cannot decide what is worth solving.

The objective function is a political choice. Every algorithm embeds values in its optimization target; the values disappear when the algorithm presents itself as mathematics.

The digital unconscious enables psychopolitics at scale. Algorithmic governance requires legibility of the population; legibility requires the production of behavioral data the subject does not control.

Democracy degenerates quietly. Infocracy does not overthrow democracy. It hollows it out by replacing deliberation with management, one technical decision at a time.

Debates & Critiques

Defenders of algorithmic governance argue that algorithms merely extend the human capacity for coherent decision-making — that they reduce bias, improve consistency, and make governance more responsive to complex realities. Han's response is that these benefits are real and that the costs are real too, and that the costs are specifically the elimination of the space where human beings, as political beings, negotiate what they want to become. The benefits of optimization are measurable; the costs of eliminating the political are not. The asymmetry of measurability is itself part of the mechanism through which infocracy advances.

Appears in the Orange Pill Cycle

Han Byung-Chul — On AI