The Geoeconomics of Generative AI Development Race: How Emerging Technologies Can Become National Power

As of 2025, countries around the world—most notably the United States and China—are engaged in intense competition over the development of generative AI. The race to develop cutting-edge generative AI models is no longer a matter of technological advancement alone; it has become a central arena in the contest for national power. Unlike nuclear and space technologies development during the Cold War, generative AI is not exclusively a state-led effort. Instead, it is driven primarily by private companies, with governments, universities, and investors all participating in a complex innovation ecosystem. This represents one of the defining features of today’s emerging technologies.

This report does not evaluate the technical advancement of generative AI in each country or judge individual national policies. Instead, it analyzes the AI development race through the lens of national power—examining major powers’ strategies and how they attempt to translate technological capability into geopolitical influence. Focusing on the United States, China, Europe, and Japan, it explores how generative AI can be transformed into an instrument of national power and what strategic implications arise for Japan.

Chapter 1: The United States — Maximizing a Private Sector-led Ecosystem

The United States leads the global development of generative AI with an unparalleled innovation ecosystem encompassing foundation models, computational resources, talent, and investment. Tech giants such as Microsoft, Google, and Amazon dominate every stage from R&D to service delivery, supported by massive financial and infrastructural power. While the government plays a relatively limited role, changes of administration have led to major shifts in regulation and risk-management policy. The Department of Defense is rapidly advancing military applications of generative AI; however, the U.S. still lacks a coherent model for converting technological superiority into sustainable national power.

Chapter 2: China — The “Success” and Dilemmas of an Authoritarian Approach

China pursues a state-driven strategy under the banner of technological self-reliance and self-strengthening. Its generative-AI ecosystem is characterized by strong government control and a rapidly expanding base of AI firms. While China continues to lag behind the United States in certain areas, including advanced semiconductors, China is cultivating a competitive domestic research community. The emergence of cost-efficient models such as DeepSeek reflects this progress. Yet, Beijing’s early move toward strict regulation and legal control of generative AI has created a dilemma: balancing continued innovation with centralized management to block foreign ideological influences.

Chapter 3: Europe — Can Regulation Become Power?

Europe does not match the United States and China in foundational model development and R&D capacity, but it has taken the global lead in AI regulation most comprehensively and systematically. The EU AI Act, which entered into force in August 2024, aims to safeguard European values such as human rights and privacy while using stringent rules to defend digital sovereignty and project regulatory power abroad. However, the very strictness of these regulations risks stifling innovation within the EU. Europe is continuing to seek a sustainable balance between control and creativity, exploring whether regulatory power can translate into geopolitical influence.

Chapter 4: Japan — Searching for a Strategy as a Middle Power

Japan trails the U.S. and China in foundation model development but seeks to build an AI ecosystem through public-private partnership. The government supports the private sector’s innovation by providing computational infrastructure, while universities and research institutions play complementary roles. Japan is also developing rules to ensure the safety and trustworthiness of generative AI. Nonetheless, structural constraints—limited human resources, computing capacity, and investment scale—remain severe. Japan’s ability to keep up with global advances in generative AI will significantly influence its power as a sovereign state.

Chapter 5: Comparative Insights and Policy Recommendations

The comparative analysis reveals that each nation and region have distinctive approaches to developing generative AI. It indicates the competition over generative AI is not merely about technical capabilities, but about how states combine infrastructure, finance, human capital, and institutions to convert technology into nations’ power. Secondly, among these factors, institutional design—particularly regulation—offers the greatest room for policy maneuver. However, its effectiveness as a power remains questionable, particularly when these institutions lack robust industrial and technological base. Thirdly, as foundation models are trained on natural language data, they inevitably embed values and ideologies. The diffusion of such models overseas thus represents not only a technological export, but also an export of ideas—a growing dimension of power in the generative AI era.

Based on the various strategies of each country and region, the report outlines the following strategic directions for Japan:

Policy Recommendations

(1) Choosing the Path of a “Middle Power” in Generative AI

Japan should recognize the limits of pursuing leadership comparable to the U.S. or China. Instead, it should position itself as a middle power that focuses on areas where it can exert unique influence—particularly in governance, application, and safety.

(2) Shifting from the Foundation Model Race to Strength in Applications

Competing for investment resources for foundation model development is unrealistic. Japan should redirect limited resources toward application domains such as defense, manufacturing, healthcare, and education—fields tied closely to national needs—building unique competitiveness through specialization.

(3) Sustaining Participation through Minimal Investment and Infrastructure

While such a shift in focus toward applications is necessary, it should not come at the expense of completely abandoning foundation model development. A certain level of sustained investment in foundation model research, computational infrastructure, and human-resource development remains vital for Japanese R&D organizations and private firms to stay engaged in the global ecosystem. This ensures Japan’s continued participation in international standard-setting and governance discussions.

(4) Accelerating a Regulatory Approach

Japan should enhance domestic governance frameworks to ensure safety, transparency, and accountability in the realm of generative AI. When regulation is paired with effective evaluation capabilities, it can serve as a strategic asset. This approach would allow Japan to manage the risks associated with AI while also demonstrating leadership and competitiveness in promoting trustworthiness and safety within the field of AI applications.

(5) Reconstructing a Multi-Layered International Cooperation Strategy

Tokyo must pursue domestic regulation and international norm-building in tandem. “The Hiroshima AI Process” offers one model of success. Japan should deepen ties with the EU, whose regulatory orientation is similar, while continuing technical cooperation with the U.S. and expanding partnerships with resource-constrained Asian nations to build a multi-layered cooperative framework.

(Photo Credit: blackdovfx / Getty Images)

Authors