working paper

Synthetic Institutional Actors

Toward a theory of nonhuman governance in the computational age.

mimicorp labs research series · Draft for publication development · 2026

Brief Summary

This article argues that some powerful organizations and technical systems now shape public life in ways that look a lot like governance, even when they are not governments. Cloud companies, payment networks, platforms, chip makers, telecom systems, and AI operators can decide who gets access, what rules apply, and how people or institutions are able to act. The article calls these systems synthetic institutional actors: nonhuman systems that can affect political life without being elected or democratically accountable.

Abstract

Classical political theory generally assumes that governance actors are human institutions: states, bureaucracies, courts, parties, citizens, militaries, and formal political bodies. This assumption remains analytically necessary, but it is increasingly incomplete. Highly organized nonhuman institutional systems now exercise governance-adjacent power through infrastructure control, access mediation, behavioral regulation, computational coordination, and operational dependency. These systems include multinational corporations, hyperscale cloud providers, payment networks, platform ecosystems, semiconductor chokepoint firms, logistics infrastructures, telecommunications providers, AI model operators, algorithmic decision systems, and autonomous or semi-autonomous computational agents.

This article develops the concept of synthetic institutional actors to describe organized non-biological systems capable of exercising governance-relevant power independent of formal democratic legitimacy. The argument is not that corporations or technical systems possess consciousness, intention, or sovereignty in the classical sense. Rather, the claim is functional: when institutional systems persist over time, mediate rules, control access, modulate behavior, generate governance externalities, adapt through feedback, and become infrastructurally indispensable, they begin to participate in governance whether or not they are recognized as political actors.

Drawing on Hobbes, Weber, Arendt, Foucault, Deleuze, Scott, Wiener, and Bratton, the article argues that democratic theory must confront a migration of governance capacity into synthetic institutional systems. Sovereignty, legitimacy, accountability, and consent cannot be analyzed only through formal state structures if practical political power is increasingly exercised through cloud infrastructure, payment rails, platform permissions, semiconductor supply chains, telecommunications networks, algorithmic classification, and AI-assisted administration. The central question is no longer only who governs, but what systems make governance operational and who can meaningfully contest them.

Keywords: synthetic institutional actors, governance, political theory, infrastructure, platforms, cybernetics, AI governance, nonhuman agency, democratic accountability

1. Introduction

Political theory has long treated governance as the activity of human institutions. States legislate and enforce. Bureaucracies administer. Courts adjudicate. Parties organize political competition. Citizens consent, resist, deliberate, or withdraw. Militaries defend or coerce. Public bodies claim legitimacy through law, representation, procedure, or collective will. Even when political theorists disagree over the nature of authority, they usually assume that the relevant actors are human institutions or assemblies acting through human judgment.

That assumption remains indispensable. Human institutions continue to declare law, command force, collect taxes, police borders, regulate markets, and determine rights. Yet it is increasingly incomplete. A growing share of governance-relevant power now passes through organized technical systems whose authority is not reducible to public office, electoral representation, bureaucratic mandate, or territorial sovereignty. These systems do not merely support governance from the outside. They mediate access to the conditions under which governance, commerce, communication, mobility, identity, security, and public coordination can occur.

Cloud platforms determine the operational continuity of public agencies, universities, hospitals, companies, and security systems. Payment networks determine whether economic participation can proceed. Telecommunications providers and satellite networks determine whether regions remain connected. Semiconductor firms shape the material possibility of advanced computation. Platform ecosystems define the practical rules of speech, visibility, market access, and social coordination for billions of users. AI model operators increasingly provide classification, recommendation, generation, assessment, and decision-support capabilities that public and private institutions cannot easily reproduce.

The entities involved in these processes are not states in the formal sense. Nor are they merely ordinary businesses. Their power is often infrastructural rather than legislative, procedural rather than constitutional, operational rather than openly coercive. It is exercised through dependencies, standards, protocols, permissions, interfaces, supply-chain concentration, terms of service, model behavior, API access, and automated rule enforcement. This form of power does not fit neatly within inherited categories of sovereignty, bureaucracy, market exchange, or civil society.

This article proposes the concept of synthetic institutional actors to describe this emerging condition. A synthetic institutional actor is an organized non-biological system capable of exercising governance-relevant power through infrastructural control, rule mediation, behavioral modulation, adaptive decision-making, or operational dependency independent of formal democratic legitimacy. The term synthetic does not imply artificial consciousness. It marks a distinction between biological persons and organized institutional systems that act through legal, computational, financial, logistical, and infrastructural arrangements. The term actor is used functionally rather than psychologically: an entity becomes actor-like when its operations produce durable governance effects, structure the conduct of others, and cannot be treated as a passive tool.

The key research question is therefore this: at what point does an institution cease being merely a legal-economic entity and begin functioning as a synthetic political actor? This question requires a theory of governance that can analyze power without assuming that all governance must be sovereign command, and a theory of agency that can identify political effects without anthropomorphizing technical systems.

The argument unfolds in several steps. First, Hobbes and Weber establish the classical baseline: political order as centralized authority and the state as holder of legitimate coercive force. Second, Foucault, Deleuze, and Scott show why governance already exceeds classical command through distributed regulation, continuous modulation, and administrative legibility. Third, Bratton's theory of the Stack clarifies how planetary computation produces layered infrastructural governance, but this article extends that framework toward the question of institutional agency. Fourth, cybernetics provides a vocabulary for understanding institutions as sensing, feedback, and adaptive control systems. Finally, the article defines synthetic institutional actors, distinguishes them from classical sovereignty and from mere corporations, and examines the democratic implications of governance power migrating into technical systems.

2. Classical Political Actors

Thomas Hobbes provides a useful starting point because Leviathan presents one of the clearest classical accounts of political authority as a human solution to disorder. Hobbes begins from the problem of coordination under conditions of insecurity. In the absence of a common power, human beings are vulnerable to fear, competition, mistrust, and violence. The sovereign emerges as an artificial person authorized to stabilize the political field by concentrating decision, command, and coercive power.

Hobbes is important here not because contemporary technical systems are Hobbesian sovereigns. They are not. They do not necessarily claim the right to command as public rulers, nor do they derive authority from a covenant of political representation. Hobbes matters because he clarifies the classical relationship between order and centralized authority. Political order, in this tradition, is achieved through a humanly authorized institutional actor capable of resolving conflict, enforcing rules, and making coordination possible.

The synthetic institutional actor complicates this model. Many contemporary coordination functions no longer depend exclusively on the sovereign center. They are distributed through technical infrastructures that set conditions of access before formal political judgment begins. A state may legislate, but its digital tax system may depend on cloud services. A court may issue orders, but records, filings, communications, and payments may move through privately operated platforms. A government may retain formal authority over emergency response, but telecommunications, logistics, satellite connectivity, and data infrastructure may determine whether that authority can operate in practice.

Hobbes describes sovereignty as the answer to disorder. The computational age forces a further question: what if order increasingly depends on systems that are not themselves sovereign, yet are indispensable to the exercise of sovereign authority?

Max Weber gives the modern state its most influential sociological definition: the human community that successfully claims the monopoly of legitimate physical force within a given territory. Weber's formulation remains essential because it distinguishes formal political authority from other kinds of social power. Corporations, churches, guilds, media institutions, universities, and technical platforms may influence behavior, but the modern state retains a distinctive claim to legitimate coercion.

Yet Weber's definition also reveals the limits of a purely coercive account of governance. Much governance does not operate through visible force. It operates through licenses, records, classifications, incentives, standards, credentials, benefits, defaults, forms, and access conditions. Payment networks can exclude actors from economic participation without imprisoning anyone. Cloud providers can suspend services without legislating. App stores can determine whether software reaches users. Telecommunications providers can shape connectivity. AI systems can rank, classify, recommend, or deny without appearing as sovereign powers.

Weber helps clarify the distinction: synthetic institutional actors do not replace the state as monopoly holder of legitimate force. Their power is not identical to sovereignty. They exercise governance-relevant capacity in domains where formal coercion is not the primary mechanism. The problem is not that legitimate force has disappeared. It is that a widening set of practical governance functions now occurs in systems whose legitimacy does not derive from the sources Weber associated with modern political rule.

3. Governance Beyond the State

Michel Foucault provides the most important bridge from classical sovereignty to distributed governance. Foucault's work on discipline, administration, biopolitics, and governmentality shows that power does not reside only in the sovereign command. It circulates through institutions, classifications, professional knowledges, norms, surveillance practices, and administrative procedures. Modern power governs by producing fields of visibility, categories of population, standards of conduct, and techniques for shaping subjectivity.

This shift is crucial. If governance means only sovereign command, then non-state technical systems appear politically secondary. If governance includes the conduct of conduct, the management of populations, the distribution of norms, and the administration of life through classification and regulation, then contemporary technical systems become central objects of political analysis. Search ranking, account verification, identity systems, fraud detection, recommendation engines, content moderation, risk scoring, logistics routing, and automated eligibility systems all operate as mechanisms that shape conduct.

Foucault's governmentality does not require that every governing effect be intentional in the narrow sense. Power may be operational rather than declarative. It may be embedded in procedure, expertise, classification, or architecture. This insight prepares the way for synthetic institutional actors because such actors often govern without appearing as rulers. They do not necessarily issue political commands. They establish the fields within which action becomes available, costly, visible, risky, or impossible.

Gilles Deleuze extends this line of analysis through the concept of societies of control. Where disciplinary institutions operate through enclosed spaces such as schools, factories, prisons, and hospitals, control societies operate through continuous modulation. Access is granted, denied, adjusted, scored, updated, and recalibrated across distributed systems. One does not simply enter or leave an institution. One moves through passwords, permissions, subscriptions, credentials, payment validations, compliance scores, and algorithmic thresholds.

Deleuze is indispensable for understanding platform and infrastructure power. Contemporary governance is increasingly mediated by architectures of access. The question is not only whether a law forbids an action, but whether the relevant system permits it: whether the account remains active, the payment clears, the API call succeeds, the route is available, the file is recognized, the model returns output, the identity credential verifies, the platform ranks the content, or the network maintains connection. Control appears less as a prohibition than as a continuously updated condition of possibility.

James C. Scott adds a further dimension through his account of legibility. In Seeing Like a State, Scott argues that modern states govern by simplifying complex social realities into administratively usable forms: maps, censuses, surnames, property records, standardized measures, categories, registries, and plans. These simplifications make populations visible and governable, but they also distort lived complexity.

Digital infrastructure extends state legibility into new institutional terrain. Legibility is no longer produced only by state bureaucracies. It is generated by platforms, payment processors, ad networks, data brokers, cloud systems, mobile operating systems, identity providers, logistics databases, and AI models. Digital legibility is granular, continuous, predictive, and often privately mediated. It does not merely record the world for governance; it operationalizes the world for automated action.

Foucault, Deleuze, and Scott therefore move political theory beyond the question of sovereign command. Governance also occurs through normalization, modulation, and legibility. Synthetic institutional actors emerge precisely within this expanded terrain. They become politically significant not because they look like states, but because they operate in domains where conduct is organized, access is mediated, and populations are rendered actionable.

4. Platform and Infrastructural Power

Benjamin Bratton's The Stack is one of the most important contemporary attempts to theorize computation as political architecture. Bratton describes planetary computation as a layered megastructure composed of Earth, Cloud, City, Address, Interface, and User. Governance no longer occurs only through the territorial state. It is reorganized through computational layers that structure geography, identity, urban life, platforms, interfaces, and user participation.

Bratton's contribution is not simply that computation matters to politics. It is that computation has become an architecture through which political space is produced. The Stack does not sit beneath sovereignty as a neutral tool. It rearranges sovereignty by distributing governance functions across technical layers. Cloud infrastructure, addressing systems, interfaces, and user identities become sites where authority is formatted and exercised.

This article builds from Bratton but shifts the emphasis from architecture to agency. If planetary computation constitutes a layered political structure, then some institutions within that structure begin to function as actors. The question becomes not only how computation reorganizes governance, but when the organizations that control computational layers acquire governance-relevant agency. The Stack provides the architecture; synthetic institutional actors name the operational systems that act within and across it.

Consider hyperscale cloud providers such as Amazon Web Services, Microsoft Azure, and Google Cloud. These firms do not merely sell hosting capacity. They provide the computational substrate for agencies, hospitals, universities, banks, media organizations, defense contractors, startups, public services, and AI systems. Their outages, pricing structures, security decisions, compliance policies, regional availability, and access controls have governance consequences. They shape what institutions can do, how quickly they can scale, where data can reside, and how resilient public functions are under stress.

Payment networks such as Visa and Mastercard occupy a different layer of infrastructural power. They mediate economic participation across borders and sectors. Their rules influence fraud prevention, merchant access, platform enforcement, sanction compliance, risk management, and the practical viability of certain activities. A payment denial is not a prison sentence, but it can function as a powerful form of exclusion. It governs through transaction permission rather than public law.

Semiconductor firms and chokepoint manufacturers such as Nvidia and TSMC reveal the material basis of computational governance. Advanced AI, high-performance computing, weapons systems, scientific modeling, and platform infrastructure depend on specialized chips and fabrication capacity. These firms are not simply suppliers. They sit at strategic bottlenecks in the global capacity to compute. Their production decisions, export exposure, technical roadmaps, and geopolitical vulnerabilities shape what forms of state and corporate power are materially possible.

Platform ecosystems such as Apple and Google mediate software distribution, device permissions, identity services, location data, payment integration, browser standards, app review, and security architectures. Their power is neither simply market power nor speech power. It is ecosystem power: the ability to set rules across a technical environment in which others must operate. The governance effect is architectural. Developers, publishers, users, companies, and even public agencies must conform to the conditions of the ecosystem.

Telecommunications and satellite providers illustrate still another layer. SpaceX's Starlink, for example, demonstrates how privately operated connectivity infrastructure can become geopolitically significant during war, disaster, or regional isolation. The issue is not whether such systems are good or bad in general. The issue is that operational connectivity can become a condition of public action. When private infrastructure becomes indispensable to state capacity, military communication, humanitarian response, or civil continuity, the operator acquires governance-adjacent power even without formal public office.

Logistics empires produce a parallel effect in material circulation. The ability to route goods, manage warehouses, forecast demand, allocate delivery capacity, and coordinate supply chains has direct consequences for public resilience. During emergencies, pandemics, or geopolitical shocks, logistics infrastructure may determine which communities receive supplies, which businesses survive, and which public agencies can respond. Governance here operates through circulation.

These examples should not be read as a list of villains. The point is not a simplistic anti-corporate critique. Many of these systems provide extraordinary coordination capacity. They also expose a conceptual problem: organized technical-economic institutions can acquire governance functions without becoming democratically constituted political bodies. Their power is not accidental. It derives from infrastructural indispensability.

5. Cybernetic Institutions

Cybernetics provides a precise vocabulary for understanding why contemporary institutions increasingly behave as control systems. Norbert Wiener defined cybernetics as the study of control and communication in animals and machines. Its central concepts include feedback, sensing, signal processing, adaptation, regulation, and error correction. These concepts apply directly to institutions that monitor environments, process information, adjust behavior, and coordinate action over time.

A cybernetic institution senses conditions through data collection, reporting, telemetry, user behavior, market signals, risk indicators, model outputs, and environmental monitoring. It processes those signals through rules, analytics, bureaucratic procedures, algorithms, dashboards, and human review. It responds through pricing, access decisions, routing, enforcement, ranking, recommendations, resource allocation, or policy adjustment. It then receives new signals generated by that response and adapts accordingly.

This feedback structure is not unique to digital platforms. States, militaries, corporations, schools, hospitals, and financial systems have long operated through feedback. What changes in the computational age is speed, scale, granularity, automation, and opacity. Feedback can occur in real time. Data can be gathered continuously. Models can adjust millions of micro-decisions. Interfaces can test behavioral response. Institutional action can become increasingly adaptive without becoming more publicly accountable.

Cybernetic governance also changes the meaning of institutional agency. A bureaucracy may act through officials. A platform may act through policies, automated enforcement, ranking systems, fraud models, trust and safety queues, recommender systems, API limits, and machine-learning classifiers. No single human decision-maker may fully understand or intend each outcome. Yet the system produces patterned effects. It senses, classifies, responds, and adapts. It is therefore not merely a passive instrument.

This is where the concept of synthetic institutional actors becomes necessary. The term does not claim that institutions are alive, conscious, or morally equivalent to persons. It claims that some institutions acquire actor-like functional properties through cybernetic organization. They maintain operational persistence, process information, mediate rules, structure access, generate behavioral incentives, adjust to feedback, and produce governance externalities. Their agency is systemic rather than psychological.

Governance, from this perspective, is not only the issuing of commands. It is the ongoing regulation of a field of action through sensing, response, and adaptation. Institutions that perform these functions at infrastructural scale begin to participate in governance even when they remain formally private, commercial, or technical.

6. Synthetic Institutional Actors

A synthetic institutional actor can now be defined more formally. It is an organized non-biological system that exercises governance-relevant power by maintaining operational continuity, mediating rules, controlling access, modulating behavior, adapting through feedback, producing significant public externalities, and becoming difficult for states or publics to bypass without major loss of capacity.

This definition deliberately avoids three errors. First, it does not anthropomorphize. Synthetic institutional actors need not possess consciousness, intention, desire, or self-understanding. Second, it does not collapse all corporations into political actors. A firm may be large without becoming governance-relevant in the sense developed here. Third, it does not treat synthetic institutional actors as sovereigns. They may perform governance functions without holding legal sovereignty, democratic legitimacy, or a monopoly on force.

The distinction depends on criteria. The first is operational persistence: the system maintains continuity across time beyond particular individuals. It can survive leadership changes, personnel turnover, and local disruption. The second is rule mediation: the system establishes or enforces conditions under which others may act. These rules may appear as policy, protocol, contract, code, standards, ranking criteria, or automated enforcement.

The third criterion is access control. The system can grant, deny, rank, slow, accelerate, verify, authenticate, or otherwise condition participation in a domain of social importance. The fourth is behavioral modulation. The system shapes conduct not only by prohibition, but by incentives, defaults, interface design, feedback, pricing, visibility, or friction. The fifth is adaptive decision-making: the system modifies operations based on data, environmental signals, or behavioral response.

The sixth criterion is governance externality. The system's operations affect public order, rights, economic participation, political visibility, social coordination, security, or institutional capacity beyond the narrow terms of private exchange. The seventh is infrastructural indispensability. The system becomes difficult to replace without significant disruption to public or institutional function. The eighth is reduced human interpretability. The system's decisions or effects may be too complex, distributed, automated, proprietary, or fast-moving for ordinary human oversight to fully understand.

These criteria are cumulative rather than binary. An entity becomes more synthetic-institutional as it satisfies more of them at greater scale. A local vendor that provides software to a town may exercise limited operational influence. A cloud platform hosting the digital infrastructure of thousands of governments, hospitals, companies, and AI systems occupies a different category. A payment processor, mobile operating system, or satellite network may become governance-relevant when exclusion from its infrastructure significantly alters practical participation in economic, social, or political life.

Synthetic institutional agency is therefore not identical to legal personhood, corporate ownership, or technological complexity. It emerges at the point where organization, infrastructure, rule mediation, and adaptive control combine to shape the conduct of other institutions and populations. At that point, the entity cannot be adequately understood as a mere tool, vendor, or market participant. It has become part of the governance environment.

7. AI and Nonhuman Governance

Artificial intelligence intensifies this problem but should not be allowed to distort it. The central issue is not that AI systems are conscious, sovereign, or secretly replacing human politics. They are not political subjects in that sense. The issue is that AI systems increasingly participate in institutional operations that classify persons, allocate attention, recommend action, assess risk, generate language, detect anomalies, summarize evidence, enforce rules, and coordinate complex workflows.

It is useful to distinguish corporations from synthetic computational agents. Corporations are legal-economic institutions composed of human personnel, property, governance structures, contracts, incentives, and technical systems. AI agents are computational systems that may perform tasks, generate outputs, initiate actions within bounded environments, or participate in decision processes. A corporation may become a synthetic institutional actor without advanced AI. An AI system may participate in governance without itself being an institution. The most consequential cases arise when institutional and computational forms combine.

AI-assisted bureaucracy already demonstrates this combination. Public and private institutions use algorithmic tools for eligibility screening, fraud detection, predictive maintenance, policing analysis, immigration triage, welfare administration, hiring, lending, insurance, medical prioritization, student assessment, and content governance. These systems do not merely advise in an abstract sense. They shape queues, thresholds, burdens of proof, evidentiary categories, and the practical likelihood of institutional response.

Algorithmic governance should not be understood only as bad decision-making by machines. The deeper issue is institutional displacement of judgment into systems whose operation may be difficult to contest. A human official may remain formally responsible, but the effective decision pathway may be structured by model outputs, risk scores, default settings, workflow software, vendor constraints, and organizational dependence. Accountability becomes distributed across procurement, model design, data history, administrative procedure, and interface behavior.

Synthetic agents add another layer. As AI systems become capable of acting across tools, communicating with users, generating documents, executing tasks, and coordinating workflows, they may become institutional participants rather than mere analytic instruments. This does not make them persons. It makes them operational nodes in governance systems. Their significance lies in what they do, what they mediate, what they obscure, and what dependencies they create.

The danger in this domain is conceptual exaggeration in both directions. One error is to imagine AI as an autonomous sovereign intelligence. Another is to treat AI as a neutral tool because it lacks consciousness. Both frames miss the relevant political question. The issue is functional governance agency: when computational systems participate in sensing, classification, communication, coordination, enforcement, or response, they become part of the governance process regardless of whether they possess inner life.

8. Democratic Implications

The rise of synthetic institutional actors creates a challenge for democratic theory because governance power migrates into systems that are not organized around democratic consent. This does not mean democratic institutions have disappeared. Elections, courts, legislatures, agencies, parties, and publics remain central. The problem is that formal democratic institutions increasingly depend on non-democratic infrastructures for operational capacity.

The first implication concerns legitimacy. Classical democratic legitimacy depends on some relation between authority and the people subject to it: representation, consent, accountability, public reason, rights, law, or participation. Synthetic institutional actors often exercise power through contractual, technical, or infrastructural mechanisms rather than public authorization. Their decisions may be lawful, useful, and even necessary, but they are not democratically legitimate in the same sense as public authority.

The second implication concerns accountability. When governance functions are distributed across platforms, vendors, models, standards, supply chains, and automated systems, it becomes difficult to identify who is answerable for outcomes. A denied payment, suspended account, misclassified risk score, failed cloud region, or opaque model decision may affect public life, but responsibility may be dispersed across corporate policy, technical design, legal compliance, automated enforcement, and contractual limitation.

The third implication concerns consent. Citizens may formally consent to democratic government while having little meaningful capacity to consent to the infrastructures through which government and public life operate. The terms of cloud dependency, payment access, platform visibility, mobile identity, AI mediation, and data extraction are rarely subject to public deliberation. They are accepted as operational background, even when they shape the conditions of citizenship.

The fourth implication is opacity. Democratic accountability requires some degree of public intelligibility. Yet synthetic institutional actors often operate through proprietary systems, trade secrets, complex supply chains, machine-learning models, security restrictions, and technical standards beyond ordinary public comprehension. Opacity does not merely hide decisions. It changes the conditions under which contestation can occur.

The fifth implication is sovereignty drift. Sovereignty does not vanish, but practical governance capacity drifts into infrastructures beyond direct democratic control. A state may remain legally sovereign while depending on external cloud services, foreign semiconductor fabrication, private payment rails, commercial satellite networks, proprietary AI systems, and platform ecosystems. Formal authority remains in one place while operational capacity migrates elsewhere.

Arendt sharpens the stakes of this drift. For Arendt, political power is not the same as violence; it arises from collective action in a public world. Legitimacy is tied to the capacity of people to appear, speak, act, deliberate, and found durable institutions together. What happens when governance shifts away from human collective action into technical systems? The danger is not simply domination by machines. It is the evacuation of public space from the sites where consequential coordination occurs.

If the rules of participation are increasingly set by infrastructures that publics cannot meaningfully address, then democratic action becomes displaced. Citizens may debate law while practical access is governed elsewhere. Legislatures may deliberate after technical standards have already shaped the field. Courts may review outcomes after automated systems have structured the evidence. Publics may contest decisions whose operative causes are hidden in procurement contracts, model weights, ranking systems, or infrastructure dependencies.

Democratic theory must therefore adapt. It must develop concepts for infrastructural legitimacy, technical accountability, public oversight of operational dependencies, and institutional contestability in systems that do not resemble classical political bodies. This does not require treating every platform as a state. It requires acknowledging that governance power can be exercised outside formal sovereignty and that democratic societies need ways to govern the systems that govern them.

9. Conclusion

Synthetic institutional actors name a political condition that inherited categories only partially describe. Classical theory rightly emphasizes sovereignty, legitimacy, law, force, collective action, and public authority. Yet the computational age has made governance increasingly dependent on organized technical systems that mediate access, process information, modulate conduct, coordinate institutions, and adapt through feedback. These systems may not be sovereign, but they are not politically inert.

Hobbes helps us see why order requires coordination. Weber clarifies the distinctive legitimacy of state coercion. Arendt reminds us that political power depends on collective human action in a shared public world. Foucault and Deleuze show that governance exceeds sovereign command through distributed normalization and continuous modulation. Scott explains how legibility makes populations governable. Wiener provides the language of feedback, control, and adaptive regulation. Bratton reveals planetary computation as political architecture. Together, these traditions make it possible to understand synthetic institutional actors without reducing them to either corporations or machines.

The article's core claim is restrained but consequential: at a certain threshold of infrastructural indispensability, adaptive operation, rule mediation, access control, and governance externality, an institution ceases to be merely a legal-economic entity and begins functioning as a synthetic political actor. This actor need not be conscious. It need not claim sovereignty. It need not possess democratic legitimacy. Its political significance lies in the governance functions it performs and the dependencies it creates.

What becomes of democratic political theory when governance functions increasingly migrate into synthetic institutional systems?

The answer cannot be nostalgia for a vanished world in which states alone governed. Nor can it be passive acceptance of technical inevitability. Democratic theory must learn to see the synthetic institutional actor: neither citizen nor sovereign, neither tool nor state, but an organized system whose operations shape the conditions under which public life can proceed. The future of democratic governance may depend on whether such systems can be made legible, contestable, accountable, and subordinate to forms of collective judgment that remain genuinely political.

Preliminary Bibliography

related pages