•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Internet is often described as a borderless space where information flows freely and power is dispersed. In practice, however, the network depends on physical systems and concentrated capabilities. Beneath the ocean, hundreds of thousands of kilometers of fiber optic cables operate as key arteries of the digital economy, largely along routes shaped by telecommunications infrastructure built more than a century ago.
What has changed is not the geography of connectivity, but the subject of power. Today, major technology firms—Google, Microsoft, Meta, Alibaba, and Amazon—often referred to as Big Tech, invest not only in digital platforms but also directly in infrastructure, data, and computing power. These elements are increasingly central to artificial intelligence (AI), giving firms leverage not through traditional territorial control, but through control of the “invisible infrastructure” on which the modern economy relies.
The new power can be understood as four stacked layers that reinforce one another.
Contrary to common imagery, the internet is not an immaterial cloud. It relies on submarine fiber optic cables, landing stations, and large data centers. Over the past two decades, Big Tech has increasingly built and owned fiber routes rather than only leasing them from telecom operators. This can reduce costs and also provide direct control over the “path” that data takes—an advantage as global traffic grows.
AI—especially large models—requires substantial computing resources, typically delivered through cloud computing systems. This capacity is highly centralized and mainly held by a small number of firms. As a result, access to compute is no longer automatic; it becomes a strategic resource similar to energy in earlier industrial economies. Control of computing capability can therefore shape the pace and direction of the AI race.
Alongside traditional natural resources, the 21st century has elevated data as a key resource. Behavioral data, transactional data, and linguistic data are collected at scale by platforms such as Meta Platforms and TikTok. This information is valuable commercially and also serves as essential input for training AI models.
The final layer is AI itself, where power becomes more refined. Large language models and machine learning systems do not only process data; they “interpret” the world on behalf of humans. They influence how information is searched, organized, and presented. When a search tool or AI assistant answers a question, it does more than deliver data—it helps shape perception, representing a form of “cognitive power.”
Combined, these four layers create a new power structure. Unlike historical empires, Big Tech does not need to control territory or run a bureaucratic apparatus. Its power is infrastructural: it is widely available, yet not easily substitutable. It is also foundational, with economic, social, and political activity increasingly depending on it.
The idea of Big Tech as an “empire without territory” should be treated as a metaphor rather than a literal description. Unlike historical empires, these firms do not possess legal sovereignty, do not maintain armies, and remain subject to state regulation. Debates in the United States and Europe about antitrust, data protection, and AI governance indicate that governments continue to exert significant influence.
At the same time, the relationship between Big Tech and the state is not simply adversarial. Technology firms depend on legal, market, and policy frameworks provided by governments. Conversely, governments rely on infrastructure and technologies supplied by the private sector, particularly in areas such as cloud computing and cybersecurity. This creates a bidirectional dependence in which power is shared, though not fully transparent.
In this context, the central question is how this new form of power will be governed. If digital infrastructure and AI become the backbone of the economy and society, establishing principles of transparency, competition, and accountability becomes more urgent. The issue is therefore not only technological, but also institutional.
For developing economies, dependence on external infrastructure and platforms may bring short-term efficiency and cost benefits, but it can also create long-term risks to digital sovereignty. Participating in the global digital ecosystem without losing autonomy is described as a nontrivial challenge requiring strategic vision and policy capacity.
A common thread between the 19th and 21st centuries is not the form of power, but its logic. In the past, controlling sea routes meant controlling trade and information. Today, controlling digital infrastructure, data, and AI capabilities can confer a privileged position in the global order. While the internet may appear to flatten the surface of the world, power concentrates around strategic chokepoints embedded in the invisible infrastructures used daily.
In that environment, a “territory-less empire” is presented as a cautious but necessary way to understand how power shifts in the AI era.
Premium gym chains are entering a “golden era” that is ending or already in decline, as rising operating costs collide with shifting consumer preferences toward more flexible, community-based ways to exercise. Long-term memberships are shrinking, margins are pressured by higher rents and facility expenses, and competition from smaller, more personalized…
