Document Type : Research Paper
Authors
1
PhD, Department of International Relations, Faculty of Law and Political Science, Allameh Tabataba’i University, Tehran, Iran.
2
Associate Professor, Department of International Relations, Faculty of Law and Political Science, Allameh Tabataba’i University, Tehran, Iran.
Abstract
Introduction
Artificial intelligence (AI) has become central to the world economy. Since the deep learning revolution of 2012, followed by large language models in 2017 and the rapid spread of generative models, advances in computing and machine learning have changed how wealth and power are produced. Chips, computing hardware and data are now strategic assets.
The impact of technology on international political economy (IPE) has been widely studied, but the environment of agency that AI creates, and the pattern of relations among actors within it, remain only partly understood. Work on critical junctures looks for crises mainly in classical economic and political settings. Work on the plurality of actors still treats the state largely as a traditional institution. Studies of data politics in China rest on one case and a governance lens. Edited volumes on the IPE of AI lack a unified theory, and standard textbooks, rooted in the globalization of the 1990s and 2000s, say little about AI. In most of this scholarship the state is a regulator or traditional actor, rarely one that technology itself is reshaping.
The gap has two sides: the impact of AI on the state, and on the transnational actors entangled with it, namely international organizations, multinational corporations and other transnational institutions.
Research Aim
This study analyzes how AI alters the capacities and function of actors in IPE, above all the state and transnational actors. The central question is how AI can change the nature of actors in IPE, particularly the state? The hypothesis is narrower: AI has created a new environment of cooperation and of intertwined, complex relations among actors, and has changed their function rather than their nature, by reshaping their capacities for decision-making, data-driven operation and autonomy. A further aim is practical: to help fill the research gap and inform policymakers.
Methodology
The research is descriptive-analytical. It compares developments in IPE before and after 2012, the year of the deep learning revolution, taken as the turning point in AI development, in two before-and-after tables, one for the state and one for transnational actors. Evidence comes from scholarly literature and from reports of the WTO, the IMF, the World Bank, the Stanford AI Index and McKinsey.
The theoretical frame is complex interdependence, developed by Robert Keohane and Joseph Nye, which admits state and non-state actors and their many channels of contact into a single analysis. Four indicators derived from the theory structure the inquiry: the impact of AI on the nature of actors; the emergence of new actors; changes in goals and interests; and the pattern of relations among actors. The design is macro-level and theoretical, without case studies.
Findings
Nature, capacities and function. For the state, AI is changing governance, economic policy and state-society relations. Decisions draw increasingly on algorithmic analysis and large datasets, and in trade, supply chains and labor markets routine data-heavy tasks pass to automated systems while managers move toward supervisory and strategic roles. States that exploit AI effectively gain an advantage but also come to depend on it. Transnational actors gain too: AI widens their reach, sharpens their data analysis and strengthens their bargaining position, so states no longer dominate economic governance as fully as before. TSMC's place in the semiconductor chain shows how one firm can influence US-China tensions. In both groups, what changes is what actors can do and how they do it; their nature persists.
New actors. Two kinds emerge. Some are genuinely new: algorithmic trading systems, AI-based platforms and fintech intermediaries that shape capital markets and resource allocation. Algorithmic traders act with speed and opacity, yet their power is concentrated in firms such as BlackRock and Citadel that own the algorithms and infrastructure. Others existed before and were empowered by AI, notably cloud companies such as Amazon Web Services, Microsoft Azure and Google Cloud, whose infrastructure services, data and computing capacity create dependence and give them growing political weight.
Goals and interests. Technological capacity is joining, and in some respects displacing, resource endowments such as oil and land as the basis of competitive advantage. States and firms therefore compete for AI capabilities as a route to economic and political standing. Research on "AI-tocracy" suggests that AI can also serve political goals: investment in facial recognition in China has been linked to the its governance. In trade, interests are being redefined around data flows, digital trade agreements and AI-related trade policy.
Relations. Three features stand out. First, national actors are partly decentralized, as transnational firms gain power and states grow dependent on them for data, infrastructure and AI governance. Second, the dependence is unequal: actors with more resources and technical knowledge benefit more, new imbalances of power follow, and emerging informal regimes strain the traditional international order. Third, cooperation and competition rise together: states and firms compete over trade, productivity and technology, yet must cooperate on cybersecurity and ethical standards. The result is a pattern of intertwined, complex relations in which a shift in one domain quickly reaches others.
Conclusion
The findings support the hypothesis. AI has not turned the state or the transnational actor into something else, it has changed what they can do, how they decide and with whom they are tied. The state remains a principal actor but operates alongside platforms, cloud providers and algorithmic systems that share in generating wealth and power. The resulting pattern of intertwined and complex relations points to an international political economy moving from linear, state-to-state interaction toward a multilevel network.
Theoretically, the study extends complex interdependence by showing that AI raises both the level and intensity of interdependence and makes transnational actors more influential. It also presses realism and liberalism to account for AI and new actors, since neither, fully explains the present environment. For policy, states and international organizations cannot rely on traditional instruments of power alone. They need stronger data and technological capacity, flexible and coordinated rules for AI and data governance, and ways to anticipate the imbalances that emerging technologies create.
The study has limits. It is macro-theoretical, has no case studies, gives little attention to related technologies such as big data, data centers, quantum AI, blockchain and the Internet of Things, and does not examine regional patterns. Future research could examine individual actors through case studies and build models and theory from the themes opened here.
Keywords
Subjects
