RESEARCH · WORKING PAPER SERIES
Working Paper Series
Research on Intelligent Capitalism, Human Sovereignty, AI Governance, and the Institutional Architecture of the AI Economy.
The series circulates original, dated, citable, and versioned research before or alongside formal journal peer review. Working papers are not peer-reviewed journal publications and may be revised as research develops.
UGI-WP-2026-01
Intelligent Capitalism
UCAISM as the Third Architecture Beyond Smith and Marx
Chang-Gyu Kwag · UCAISM Global Initiative · Version 1.0 · September 2026
Working Paper / Preprint — A revised version of this research is under review at the Cambridge Journal of Economics
Platform capitalism raises analytical, political, and moral questions that neither economics nor sociology alone fully resolves. This paper argues that platform rent persists because redress arrives too slowly relative to the speed of extraction — a mechanism formalized here as the Extraction Gap. A minimal formal model shows that sustainable rent falls once an AI system negotiates on an individual's behalf ex ante, rather than contesting terms afterward. Building on the technofeudalism and digital-empires literatures, the paper develops the User-Centric AI Individual Sovereignty Model (UCAISM), combining individual ownership of data and reputation with cooperative governance of the verifying infrastructure. The argument is grounded in Kantian autonomy, Rousseauian publicness, and a Rawlsian distributive test, and evaluated against four precedents: India's UPI, Brazil's Pix, Estonia's X-Road, and Finland's basic-income experiment. UCAISM is offered as a candidate institutional architecture, not a complete equilibrium theory.
Research role: Political-economic mechanism and formalization
UGI-WP-2026-02
UCAISM: Intelligent Capitalism and the Third Architecture
Human Sovereignty and the Reconstruction of Political Economy in the Age of Artificial Intelligence
Chang-Gyu Kwag · UCAISM Global Initiative · Version 1.0 · September 2026
Working Paper / Preprint — Not peer reviewed
Artificial intelligence is transforming capitalism not only by increasing productive capacity but also by restructuring the institutional channels through which individuals acquire information, enter markets, accumulate reputation, make decisions, and exercise economic agency. Conventional debates framed primarily as capitalism versus socialism, market versus state, or growth versus redistribution do not fully capture this transformation. In AI-mediated economies, a central political-economic question increasingly concerns control over the infrastructures of agency: data, identity, reputation, market access, algorithmic intermediation, and artificial-intelligence agents. This paper develops UCAISM, or Intelligent Capitalism, as a sovereignty-centered political-economic framework for the age of artificial intelligence. UCAISM does not propose a conventional Third Way located between capitalism and socialism. It proposes a Third Architecture: a multidimensional institutional design intended to preserve decentralized market discovery, private initiative, and technological innovation while expanding effective Human Sovereignty. The framework begins with an SM-K formulation—Smith, Marx, and Kwag. Smith contributes decentralized exchange and market coordination; Marx highlights concentration and structural power; UCAISM introduces a third question: how should effective sovereignty be distributed when intelligent digital infrastructures increasingly mediate economic action? The framework is situated in relation to Hayek's dispersed knowledge, Rawls's basic structure, Sen's capability approach, Ostrom's polycentric governance, Polanyi's embeddedness, contemporary digital competition policy, and emerging fiduciary-AI research. The paper develops six principal theoretical contributions: the Sovereignty Gap, Distributed Agency, Portable Sovereignty, Universal Basic Sovereignty, Agent-Centric Intelligent Markets, and the Third Architecture. These are implemented through seven institutional pillars encompassing Human Sovereignty, portability, fiduciary agency, basic sovereignty, Growth with Equity and shared value, coopetition, and polycentric deliberative governance. The central proposition is that technological progress should be evaluated not only by productivity, innovation, income, or aggregate welfare, but also by whether institutions leave human beings more capable of forming intentions, pursuing goals, revising choices, challenging consequential decisions, employing trustworthy representatives, moving between systems, and remaining authors of their own lives.
Research role: Foundational / umbrella framework
UGI-WP-2026-03
The Agency Gap: Governing Personal AI Agents
Delegated Authority, Capability Portability, and Individual Sovereignty
Chang-Gyu Kwag · UCAISM Global Initiative · Version 1.0 · September 2026
Working Paper / Preprint — Journal-submission preparation
Personal AI agents mark a structural transition in digital intermediation. Conventional platforms primarily organize information, attention, ranking, recommendation, matching, and market access. Increasingly capable AI agents can go further: they can communicate, negotiate, purchase, schedule, coordinate, and execute actions on behalf of individuals. This transition shifts the governance problem from the architecture of choice toward the architecture of delegated agency. This paper develops the concept of the Agency Gap: the institutional distance between the capability and authority delegated to an artificial agent and the effective control retained by the individual on whose behalf that agent acts. The framework distinguishes Agent Capability (AC), Delegated Authority (DA), and Effective Individual Control (EC). It argues that technological capability and individual sovereignty need not increase together: when capability and delegated authority grow faster than effective individual control, assistance can evolve into delegated dependence. The paper integrates adjacent literatures on agentic profiles, principal–agent governance, meaningful human control, human autonomy, interoperability, and portability. Its contribution is not a priority claim over those literatures. Rather, it shifts the unit of analysis to the institutional relationship among the individual principal, the agent, the provider, and external platforms. It connects the Agency Gap to capability portability and personal capability continuity, meaningful exit, cognitive sovereignty, fiduciary loyalty, execution provenance, and reciprocal interoperability. Meta’s 2026 Muse ecosystem is used as a contemporary stress test rather than as the evidentiary foundation of the theory. The central policy objective is capability expansion without sovereignty contraction. The paper therefore proposes a Fiduciary Agent Architecture in which increasing delegation is matched by effective institutional control. Its governing proposition is simple: the agent must be loyal to the person, not merely personalized for the person.
Research role: Personal-AI-agent governance extension