December 11, 2025 – Northeast Asia is becoming a kind of laboratory for testing various models of digital sovereignty: from service dependence to industrial autonomy. Digital independence in the region is becoming more a product of hybrid connections and distributed control than the result of a closed technological cycle, writes Yuri Kolotaev. The author is a participant of the Valdai – New Generation project.
In the context of the global race for digital leadership, many political myths have emerged about digital sovereignty and how to achieve it. For most countries, the prospect of achieving some degree of self-sufficiency in computing has become a desirable goal. This has created a niche in which large digital corporations are willing to offer their “services” in building digital sovereignty, from infrastructure provision to data localisation.
However, state policies that rely excessively on such services tend to increase external dependence rather than satisfy the demand for control over their own technological sphere. This dilemma remains relevant for both technologically weak countries and leaders, particularly those in Northeast Asia.
Today, it is no longer so easy to define what constitutes a country’s digital sovereignty. This field encompasses control over production processes (technological sovereignty), network infrastructure and information flows (data sovereignty), software and platforms (information sovereignty), and, more recently, artificial intelligence (AI sovereignty). This expansion, or even blurring, of the concept of sovereignty has resulted from the spread of state control over the digital sphere due to its high social status and critical importance for governance processes.
However, “sovereignty” has become part of the popular rhetoric not only of states but also of digital companies. Responding to demand from countries in both the West and the Global South which are seeking to develop their own technological (and not just legal) frameworks, corporations have begun offering infrastructure and cloud solutions to bridge the gap in technological inequity. This involves the creation of “sovereign clouds,” AI factories, and data processing and storage capacities as a service, which tech giants like NVIDIA, Alphabet, and Microsoft are willing to provide. As a result, the idea of state-led digital sovereignty is being appropriated and distorted by the corporate sector within the framework of digital sovereignty being offered as a “purchased service.” Most often, this entails the creation of special partnerships with states to develop digital assets localised in a specific country. However, such initiatives reflect nominal sovereignty rather than actual control over one’s own digital space, as computing algorithms, software, and, most importantly, technological components remain proprietary.
”The trend toward the spread of “sovereignty as a service” is not merely rhetorical. The key challenge facing the vast majority of countries is their limited ability to ensure digital self-sufficiency.”
The current balance in the development of cloud computing and data processing has created two key poles capable of providing the majority of core digital services: the United States and China. Their digital power is determined by the availability of infrastructural, corporate, production, and intellectual capabilities. For example, the presence of large-scale commercial cloud services capabilities, known as hyperscalers (such as Amazon Web Services and Microsoft Azure in the US, and Alibaba Cloud and Tencent Cloud in China), is crucial. Their function is to provide infrastructure (IaaS), platforms (PaaS), and software (SaaS) as services. The ability to implement related technological solutions, primarily data processing and AI, largely depends on their performance. Other countries, while capable of ensuring their independence within this framework, do so only to a limited extent.
A telling example in this context is the debate about digital sovereignty, particularly the development of “sovereign AI,” that unfolded in Northeast Asia (NEA) in 2025. The Trump administration’s aggressive positioning in the field of AI, coupled with worsening relations with China, has had a ripple effect in Japan, South Korea, and Taiwan . Fear of falling behind in the AI race and competition between the US and China for markets have led to the prioritisation of sovereign control over the AI stack in national strategies. While a region with high technological potential, NEA maintains fundamentally different initial conditions for the “sovereignisation” of the digital sphere.
In Japan, the national project to create “sovereign AI” relies heavily on partnerships with foreign corporations (Oracle, NVIDIA) through Japanese holding companies. The state encourages an “innovation-first” approach, supporting corporations as suppliers and guarantors of cutting-edge digital assets: large-scale data centres and quantum initiatives. As a result, despite significant public-private investment (approximately $135 billion by 2030), the implementation of AI solutions is largely outsourced. Computing architecture and some software solutions remain proprietary, and technological independence is replaced by regulated dependence. In other words, Japanese digital sovereignty is expressed through a model in which control is achieved not through the creation of a controlled infrastructure, but through participation in distributed technology production networks and data localisation.
South Korean society is also highly digitalised and possesses manufacturing and cloud capabilities, making it one of the few nations with real potential for an independent path in AI. The national debate over “sovereign AI” by 2025 has become one of the most heated in the region, pitting national ambitions against market pragmatism. On the one hand, the government is promoting an initiative to create an independent core AI model and a National AI Computing Centre. Its goal is to develop its own language models, trained on local data and hosted on national computing facilities. On the other hand, Korean telecom projects such as the KT Innovation Hub (a joint initiative with Microsoft) are implementing localisation on a foreign technology base, transforming the idea of AI sovereignty into a managed service. As a result, Korea is currently pursuing functional autonomy without full control over the underlying computing and algorithmic layers.
Taiwan’s AI sovereignty strategy is taking the opposite trajectory, balancing the vulnerabilities of “sovereignty as a service” with autonomy through production. Unlike Japan and South Korea, Taiwan possesses not only a political incentive for digital autonomy but also industrial capital, represented by TSMC, MediaTek, and other manufacturers that form the core of the global semiconductor industry. Taiwan exports the technological components of digital sovereignty, providing other states with access to advanced AI chips, thereby transforming control over physical infrastructure into a tool of geoeconomic influence. At the same time, there is an active public debate within Taiwan about the strategic vulnerability of this model. Control over key manufacturing technologies makes the island both a centre of gravity and a target for pressure from the US and China. In other words, the Taiwanese example demonstrates that limited independence can be compensated for by direct access to the hardware building blocks of the digital world. At the same time, Taiwan’s geopolitical vulnerability raises sovereignty issues of an entirely different nature.
”Thus, Northeast Asia is becoming a kind of laboratory for testing various models of digital sovereignty: from service dependence to industrial autonomy.”
The experiences of Japan and South Korea demonstrate that even with a high level of technological development, digital sovereignty is often associated with external dependence, in which the state retains control over the rules but not the technologies themselves. Taiwan, by contrast, is transforming the material foundation of the digital world into an instrument of influence, but in doing so, it increases its own strategic vulnerability. Taken together, these examples demonstrate that digital independence in the region is becoming more a product of hybrid connections and distributed control than the result of a closed technological cycle.
In the foreseeable future, the key dilemma for digital sovereignty will remain the choice between strategic autonomy and openness to global technological flows. A combination of these two tendencies is the most desirable strategy for states, and in this sense, “sovereignty as a service” does not necessarily lead to a loss of control. For many countries, it serves as an intermediate stage in the process of institutionalising new forms of technological governance. However, such a trajectory is possible with clear national priorities, where cooperation with a service provider is not an end in itself or postulated as the achievement of sovereignty. Only a combination of standards and competitive leverage can transform dependence on external platforms into a managed partnership. Otherwise, countries risk remaining consumers of the digital age, where “sovereignty” is sold by subscription.
