Sub-Politics

In the AI Story

Beck developed the concept in the 1990s to describe a phenomenon he observed across advanced industrial societies: the most important political decisions were being made in spaces that were not recognized as political. A pharmaceutical company's decision about which drugs to develop and at what price. A telecommunications firm's decision about network architecture and data retention. An agricultural biotech company's decision about which crops to genetically modify and release into the environment. Each decision had consequences for millions of people, but none was made through democratic deliberation or subject to meaningful democratic oversight.

In the AI industry, sub-politics operates through three categories of decision. Architectural decisions—choices about model training objectives, safety calibrations, what the system will and won't do—determine the cognitive environment for millions of users. A decision to optimize for confident, fluent responses teaches, through millions of interactions, that questions have immediate answers and uncertainty is deficiency. Interface decisions—response latency, default availability, visual design—shape behavioral patterns and boundaries. A product team's choice of sub-second response eliminates the pause where reflection occurs. Deployment decisions—where tools are introduced, at what pace, with what expectations—determine organizational conditions under which individuals experience the tool. An executive decision to reduce team size based on assumed AI productivity transfers risks previously distributed across twelve people to five.

The democratic deficit is not a bug but a feature of systems where the most consequential decisions about cognitive life are made in spaces never designed for democratic deliberation. Formal political institutions—legislatures, regulatory agencies—can intervene after tools are built, regulating outcomes. But they do not participate in design decisions that determine cognitive effects before any regulation applies. The sub-political spaces remain unaccountable because the expertise required to understand them is concentrated in the institutions that operate within them, and those institutions have no structural incentive to submit to democratic oversight that might constrain their sub-political autonomy.

The EU AI Act and American executive orders represent attempts to bring sub-political AI decisions under democratic governance, but they operate at the level of supply-side regulation—constraining what companies may build—rather than participating in the demand-side design choices that determine how tools interact with human cognition. The gap between what regulation can reach and where the most consequential decisions are made remains structural, and closing it requires institutional innovations that give democratic publics voice in sub-political spaces without paralyzing the innovation those spaces produce.

Origin

Beck introduced sub-politics in Risk Society (1986) and developed it fully in Ecological Politics in an Age of Risk (1995). The concept built on Michel Foucault's analysis of biopower—the exercise of power through management of populations—and extended it to technological systems. Where Foucault examined how institutions govern bodies through normalization, Beck examined how technological institutions govern futures through design decisions that populations experience as givens rather than choices.

The concept emerged from Beck's fieldwork observing how environmental and technological risks were actually governed—not through the legislative processes civics textbooks describe, but through informal networks of technical experts, corporate managers, and regulatory officials making consequential choices in meetings the public never sees. He called this the 'reinvention of politics'—not the end of politics but its migration to new spaces that had not been recognized as political and were therefore not subject to the accountability mechanisms democratic politics requires.

Key Ideas

Migration of Consequential Power. The most important decisions about technological futures are made not in legislatures but in corporate design meetings, research laboratories, and engineering sprints—spaces with no democratic mandate or accountability.

Velocity Asymmetry. Sub-political decisions move at the speed of product development; formal political responses move at the speed of legislation—the gap between them is not closing but widening as deployment accelerates.

Expertise as Political Power. Sub-politics concentrates power in those who understand technological systems deeply enough to make design choices—creating a new form of technocratic authority operating outside democratic oversight.

Democratic Deficit. Populations affected by sub-political decisions have no institutional mechanism for participating in those decisions—no voice, no vote, no standing—creating a governance gap that Beck identified as the central political crisis of the risk society.

The Reinvention Imperative. Bringing sub-political decisions into democratic accountability requires institutional innovations that do not yet exist—not the elimination of sub-political spaces but their integration into governance structures that give affected publics voice in the decisions that shape their lives.