We are living through a great social transformation comparable to the industrial revolution. What sets it apart is not only its magnitude, but its multidimensional convergence: technological, economic, political, and epistemic shifts unfolding simultaneously and reinforcing one another. This shift places growing strain on inherited political cleavages and ideological frameworks, many forged under very different technological and institutional conditions. Mega-trends such as digitalization, artificial intelligence, platform economies, biotechnology, and the reconfiguration of global networks are not just reshaping production and exchange; they are redefining the conditions under which governance operates. The result is a systemic transformation that penetrates the foundations of contemporary political and economic orders.
At the core of this transformation is a radical expansion of the institutional possibility frontier. Technological change does not merely enhance existing arrangements; it reshapes what is feasible, enabling new forms of coordination, monitoring, enforcement, and collective decision-making that were previously impractical or unimaginable. As a result, institutional designs once confined to theory and speculation may now become operationally relevant. What is emerging is not a simple adaptation of existing systems, but a mutation of their underlying architecture.
In this context, the central question is: what place, if any, will classical liberal principles, values, and institutional arrangements occupy within this emerging order? Put differently, to what extent do current technological and structural trends constrain, erode, or, alternatively, enable new instantiations of liberal governance along classical lines? And, reciprocally, to what extent can these trends themselves be shaped, redirected, or constitutionally framed by insights derived from the classical liberal tradition?
The issue thus becomes operational and strategic. It is not enough to ask whether liberalism will survive this transformation; the more pertinent question is what kind of effort is required to ensure that the evolving architectures of governance and economic coordination remain compatible with (if not explicitly grounded in) the ethos of individual liberty, pluralism, voluntary association, and limited government. How to embed, at the level of rules, constraints, and enabling frameworks, those principles that have historically underpinned liberal democratic orders, within the new technological and economic infrastructures that are rapidly becoming the backbone of 21st-century societies.
AI and the Problem of Governance in a Post-Industrial Order
A first step requires adopting an analytical stance that recognizes a significant discontinuity between the ideological and institutional equilibria of the 20th and 21st centuries. This helps avoid the misleading assumption that antitrust doctrines, regulatory philosophies, or deregulatory responses developed under past technological conditions can be simply extended or marginally adapted to current challenges. The inherited conceptual, legislative, and legitimizing apparatus is, at best, only partially adequate to a transformed environment.
A more realistic approach is therefore dual: it preserves and critically mobilizes the insights of the 20th century while simultaneously engaging in a more ambitious effort to rethink and redesign governance frameworks and institutional architectures; ones that remain aligned with classical liberal principles yet are structurally attuned to the opportunities and constraints of the new technological landscape.
In this respect, the governance challenges raised by the emergence of artificial intelligence as a central driver of social, economic, and cultural transformation offer a paradigmatic illustration. Focusing on these challenges allows us to get a better understanding of the concrete ways the classical liberal tradition has to respond to the new realities of the 21st century.
Much of the initial debate regarding AI has focused on the distinction between inner alignment (the coherence between an AI system’s objectives and its internal processes) and outer alignment (the compatibility of those objectives with human values and social goals). Yet it has quickly become clear that this framing, while necessary, is not sufficient. Beyond alignment lies a more fundamental issue: the evolution of the interface between artificial and human intelligence generates a cascade of institutional implications, requiring a chain-reaction reconfiguration of entire governance structures and systems.
The problem thus expands from isolated and well circumscribed regulatory concerns and policy interventions to the design of an overarching governance architecture capable of managing complex, multi-level interactions across domains. What is required are multi-level, polycentric, and functionally differentiated systems of governance, able to embed these challenges within nested and modular institutional arrangements. Such systems must accommodate the diversity of collective choice situations and coordination problems that define the emerging technological landscape, while remaining adaptive to its ongoing transformation.
An Unparalleled Challenge and the Constitutional Political Economy Response
The challenge is historically unprecedented, both in magnitude and in structural complexity. For the first time, governance systems must contend with the systematic integration of two distinct forms of intelligence and agency—human and artificial—within a unified framework of social coordination. This is not a marginal adjustment to existing divisions of labor, but a qualitative transformation that reshapes the foundations of economic production, knowledge generation, and decision-making authority. (Compounding this is the broader context of simultaneous demographic, social, and cultural shifts unfolding at a global scale, which amplify the complexity and interdependence of the problems at hand).
Hence the challenge of how to respond to these transformations while maintaining a classical liberal stance: What kind of analytical and normative framework allows us to engage the governance challenges posed by artificial intelligence in a way that preserves, or at least renders compatible, the core elements of the classical liberal ethos?
This is where constitutional political economy becomes especially relevant.
The argument advanced here is that there is currently in the classical liberal tradition a vehicle perfectly suited for this task: constitutional political economy (CPE). CPE provides an approach capable not only of bridging the gap between foundational commitments (such as individual liberty, limited government, and pluralism) and the concrete institutional designs required by the new technological environment but also to offer a coherent framework for that task enabling a shift in focus from ad hoc regulatory interventions to the design of consistent higher-order rules, constraints, and procedures that structure the interaction between human and artificial agents.
As such, it stands out not only as the most consistent instrument for a classical liberal approach to AI governance, but also as one of the most advanced analytical instruments available in general for addressing the broader challenge of articulating normative and institutional responses to technological change in a domain that remains, as yet, only partially specified and insufficiently theorized.
What is Constitutional Political Economy?
As James Buchanan (Nobel laureate in economics and a founding figure of constitutional political economy as a distinct research program) famously argued, the central analytical concern of this approach is the problem of “choice among rules,” not merely choices within them; i.e. a shift of the focus of the analysis from choices made within given institutional constraints to the choice and design of those constraints themselves. Rather than examining how individuals and organizations behave under existing rules, CPE asks questions about the formation, justification, and revision of those rules. In doing so, CPE integrates positive analysis of incentives and behavior with normative inquiry into legitimacy, consent, and mutual advantage, offering a systematic framework for understanding how governance systems can be both effective and justifiable.
Beyond its analytical reorientation, CPE embodies a distinctive normative framework grounded in methodological individualism, contractarianism, and the ideal of self-governance. Institutions are legitimate insofar as they can be justified to the individuals subject to them, not by appeal to external welfare criteria but through their acceptability as the outcome—or object—of agreement under conditions of interdependence (Buchanan, 1975; Brennan and Buchanan, 1985). In this way, constitutional political economy revives the classical tradition of political economy as a “science of a legislator,” reconnecting questions of efficiency with those of institutional design, legitimacy, and governance. Moreover, it thus embodies, in a particularly refined and systematic form, the core normative commitments of the classical liberal tradition as they have evolved into the 20th and 21st centuries (Buchanan, 1990; Vanberg, 2015).
Importantly, CPE generalizes constitutional reasoning beyond the state, treating a wide range of institutional arrangements—firms, associations, platforms, and transnational regimes—as “constitutional systems” governed by rules that define authority, decision rights, and processes of revision (Ostrom, 1997; Vanberg, 1992). Therefore, at a deeper level, CPE represents the maturation of classical liberalism into a full-fledged constitutional theory of governance adapted to complex, modern societies.
This strength derives from a deep and sophisticated line of constitutional thinking developed in the works of thinkers such as Friedrich Hayek, James Buchanan, and Vincent Ostrom. Hayek’s “Constitution of Liberty” and emphasis on dispersed knowledge and spontaneous order highlights the limits of centralized control and the need for institutional arrangements that can process complexity (Hayek, 1960; 1973–79). Buchanan’s contractarian constitutionalism reframes political economy as the study of rule selection under conditions of mutual constraint and consent (Buchanan and Tullock, 1962; Buchanan, 1990). Vincent Ostrom extends this vision into a theory of polycentric governance within a “science of the artefactual”, emphasizing the importance of multi-level, self-governing systems capable of adaptation, learning and resilience (Ostrom, 1997; 1999ab).
Why Constitutional Political Economy suits so well the Governance problems related to AI?
Taken together, these contributions form a coherent intellectual architecture whose relevance is particularly evident in contexts such as artificial intelligence, where the central challenge is not merely policy adjustment but the design of new rule-based architectures capable of structuring complex, evolving systems of collective action. What distinguishes this perspective is not only its analytical depth but its institutional imagination. Classical liberal constitutionalism, in this lineage, is not a conservative or status quo-oriented doctrine; it is a framework for thinking beyond existing institutional forms. It encourages exploration of alternative governance architectures, bold redesigns of rule systems, and the creation of new forms of coordination adapted to changing technological and social conditions.
This stands in contrast to much of the current governance discourse, which often oscillates between technocratic optimization and incremental regulatory adjustment. The classical liberal approach, by contrast, invites a more ambitious project: the design of rule-based orders grounded in principles of individual liberty, consent, and dispersed authority, yet open to innovation and experimentation in institutional form. In this sense, it is not a conformist tradition but a generative one—capable of engaging the unprecedented challenges of the AI age with conceptual clarity, normative coherence, and a willingness to question and rethink the foundations of governance itself.
In a way one may say that by emphasizing constitutional rules, mechanisms of revision, and polycentric arrangements, CPE captures the dynamic and evolutionary character of liberal orders in a way that earlier formulations did not fully articulate. It is in this sense that it may be seen as the quintessence of the classical liberal ethos in the contemporary era: a synthesis that preserves its normative core—individual liberty, consent, and limited authority—while extending it into a sophisticated framework for institutional design capable of engaging the complexity and scale of 21st-century governance challenges.
An Emerging Institutional and Constitutional Turn
A shift toward institutional and constitutional thinking in artificial intelligence governance is already underway. What began as a predominantly technical field focused on model performance and alignment is progressively evolving into a domain concerned with rules, structures, and governance architectures. AI systems operate within complex, multi-level socio-technical environments where questions of authority, coordination, accountability, and legitimacy are unavoidable. The governance challenges posed by artificial intelligence are not reducible to technical alignment problems or isolated regulatory interventions; they are increasingly recognized, at their core, as problems of institutional and constitutional design.
Emerging approaches—whether framed as “Institutional AI,” “Constitutional AI,” or forms of legal alignment—reflect a growing recognition that the core problem is not simply how systems behave, but how the rules governing their behavior are defined, justified, and revised. These developments signal a move toward higher-order design questions: Who sets the rules? According to what principles? Through what procedures can they be contested and changed? While still partial and fragmented, these efforts point toward the early stages of a broader constitutionalization process in AI governance.
In such a context, the relevance of Constitutional Political Economy is not merely theoretical or prospective; it corresponds to an endogenous evolution within the field itself. What CPE offers is the capacity to systematize, clarify, and advance this emerging trend, transforming a set of disparate initiatives into a coherent framework for the design of legitimate, adaptive, and rule-based governance architectures in the age of artificial intelligence.
Equally important, CPE provides a normative framework capable of addressing the legitimacy deficit at the heart of AI governance. Current approaches often emphasize performance, safety, or compliance, but lack clear criteria for justifying the rules that govern AI systems. CPE fills this gap by grounding legitimacy in the acceptability of rules to the individuals subject to them, thus introducing standards of consent, transparency, and accountability (Brennan and Buchanan, 1985; Ostrom 1999b; Vanberg, 2015).
This perspective is particularly relevant in a context characterized by value pluralism, global interdependence, and rapid technological change. Rather than attempting to impose a fixed set of values, CPE emphasizes the design of procedures that enable diverse actors to coordinate, contest, and revise rules over time. Its emphasis on constitutional-level choice, rule revision mechanisms, and polycentric governance structures (V. Ostrom, 1997; 1999b; E. Ostrom 2010) makes it especially well adapted to conditions of uncertainty and complexity. As such, Constitutional Political Economy is not only compatible with the demands of AI governance—it provides one of the most coherent and comprehensive frameworks currently available for addressing them in a way that is both analytically rigorous and normatively grounded.
A Constitutional Strategy for a Classical Liberal Agenda for AI Governance
In brief a convergence is emerging between the institutional–constitutional turn in AI governance and constitutional political economy (CPE). This creates an opportunity for classical liberalism to engage substantively in shaping both AI governance and the broader institutional transformations of the 21st century.
First, classical liberals should move beyond downstream debates over regulation, safety, and compliance and engage upstream with the design of rule systems themselves. AI governance should be framed as a constitutional problem in the Buchanan–Ostrom sense starting from questions such as: who sets the rules, under what constraints, and through what procedures? The CPE emphasis on the “choice among rules” places the discussion at the level where the fundamental architecture of governance is determined.
Second, CPE adds a criterion often underdeveloped in AI governance: legitimacy. Technical performance—accuracy, safety, speed, or alignment—is not sufficient. Alignment itself operates within an institutional framework that establishes goals, authority, and enforcement. A classical liberal perspective therefore asks whether the rules governing AI can be justified to those subject to them, bringing consent, transparency, and accountability into the evaluation of governance systems.
Third, CPE shifts attention from declarations of principles to their institutional specification. A genuine constitutional framework requires procedures for rule-making, conflict resolution, jurisdiction, amendment, and adaptation. This is particularly important in a rapidly changing technological environment, where static rules quickly become obsolete. Constitutional procedures for contesting and revising rules are therefore essential if AI governance is to become an adaptive and sustainable institutional order rather than a collection of normative aspirations.
Fourth, uncertainty and value pluralism make institutional diversity and experimentation especially important. Corporate, legal, cooperative, decentralized, and other governance models should be able to coexist and evolve rather than being prematurely standardized. Instead of seeking agreement on a single substantive set of values, CPE emphasizes procedures through which disagreement can be managed. This logic points naturally toward polycentric governance: multiple, overlapping centers of decision-making combining autonomy with coordination. Drawing on Ostromian insights, such arrangements offer a scalable architecture for complex AI ecosystems while limiting concentrations of power.
Finally, the conceptual vocabulary of AI governance is itself still being formed. Ideas such as meta-rules, rule hierarchies, constitutional versus operational levels, legitimacy versus performance, and mechanisms of rule revision remain unsettled. Here CPE can perform an important integrative function, connecting technical, legal, ethical, economic, and policy debates within a common conceptual framework. Classical liberals therefore have an opportunity to participate not merely in particular policy disputes, but in shaping the language through which AI governance itself is conceptualized—thereby bringing classical liberal concerns with consent, dispersed authority, institutional diversity, and constitutional constraint into the intellectual foundations of the emerging governance order.
Conclusion
To sum up, Constitutional Political Economy revives the classical understanding of political economy as a discipline fundamentally concerned with the design of institutional frameworks. In the context of the AI era, classical liberals are thus called to reclaim this role: not merely as analysts or critics, but as architects of governance systems. This requires advancing CPE not as one perspective among many, but as a meta-framework capable of organizing and integrating the field as a whole. By systematically linking technical alignment to institutional incentives, legal regulation to constitutional constraints, and ethical principles to questions of legitimacy and consent, CPE may provide the intellectual infrastructure necessary to structure AI governance debates in a coherent and comprehensive manner.
The opportunity for classical liberalism is therefore foundational. The rise of AI represents, in essence, a “constitutional moment”, a juncture at which the basic rules, structures, and principles of governance are being redefined. Properly deployed, Constitutional Political Economy enables a shift from reactive commentary to proactive institutional design. It positions classical liberalism not only to respond to change, but to shape the evolving architectures of governance and embed its core principles within the normative and institutional foundations of the 21st century.
References
Brennan, Geoffrey, and James M. Buchanan. 1985. The Reason of Rules: Constitutional Political Economy. Cambridge: Cambridge University Press.
Buchanan, James M. 1975. The Limits of Liberty: Between Anarchy and Leviathan. Chicago: University of Chicago Press.
Buchanan, James M. 1990. “The Domain of Constitutional Economics.” Constitutional Political Economy 1 (1): 1–18.
Buchanan, James M., and Gordon Tullock. 1962. The Calculus of Consent. Ann Arbor: University of Michigan Press.
Hayek, Friedrich A. 1960. The Constitution of Liberty. Chicago: University of Chicago Press.
Hayek, Friedrich A. 1973–1979. Law, Legislation and Liberty. Vols. 1–3. Chicago: University of Chicago Press.
Ostrom, Elinor. 2010. “Beyond Markets and States: Polycentric Governance of Complex Economic Systems.” American Economic Review 100 (3): 641–672.
Ostrom, Vincent. 1997. The Meaning of Democracy and the Vulnerability of Democracies. Ann Arbor: University of Michigan Press.
Ostrom, Vincent. 1999a. “Polycentricity.” In Polycentricity and Local Public Economies: Readings from the Workshop in Political Theory and Policy Analysis, edited by Michael D. McGinnis, 52–74. Ann Arbor: University of Michigan Press.
Ostrom, V. (1999b). A forgotten tradition: The constitutional level of analysis. In Polycentric governance and development: Readings from the workshop in political theory and policy analysis (Vol. 151). Ann Arbor, MI, USA: University of Michigan Press.
Vanberg, Viktor J. 1992. “Organizations as Constitutional Systems.” Constitutional Political Economy 3 (2): 223–253.
Vanberg, Viktor J. 2015. “Constitutional Political Economy.” Walter Eucken Institut Discussion Paper.
Paul Dragos Aligica is a Senior Research Fellow with the F.A. Hayek Program for Advanced Study in Philosophy, Politics, and Economics at the Mercatus Center at George Mason University, KPMG Professor of Governance at the University of Bucharest, and a Senior Nonresident Scholar with the Center for Governance and Markets at the University of Pittsburgh. In 2023, Aligica was elected a Member of The Academy of Europe – Academia Europea.
(0 COMMENTS)
