The internet was built on the premise of borderlessness. That premise is being dismantled — deliberately, systematically, and with geopolitical precision. What we are witnessing is not fragmentation by accident. It is sovereignty by design.
By Vladimir Tsakanyan, PhD · Center for Cyber Diplomacy and International Security · cybercenter.space
In December 2021, Russia conducted a controlled experiment that most of the world ignored. For several hours, its national telecommunications regulator, Roskomnadzor, tested the country’s ability to disconnect from the global internet entirely — routing all domestic traffic through state-controlled exchange points, severing the external links, and operating the Russian segment of the web, known as Runet, as a sealed, self-sufficient entity. The test was imperfect. Leaks occurred. Some foreign traffic slipped through. But the infrastructure worked well enough to demonstrate the point Moscow wanted made: the kill switch exists, and it can be thrown.
This was not an act of technical curiosity. It was a statement of doctrine.
The idea that the internet is a single, unified, borderless commons has always been partly myth. It was built on American infrastructure, governed by American-adjacent institutions, and shaped — from its earliest days — by the requirements of the US military and intelligence services. What is new is not the underlying power geometry. What is new is that governments across the political spectrum, from authoritarian states to liberal democracies, have now arrived at a shared conclusion: that digital sovereignty is a legitimate, necessary, and achievable goal. They simply disagree, profoundly, on what it means.
Three Visions, One Network — For Now
The geopolitical contest over the internet’s future can be reduced, without excessive distortion, to three competing doctrines. Each is coherent on its own terms. Each is incompatible with the others. And each is being actively operationalised.
The United States has long championed what it calls a “free and open internet” — a model rooted in the conviction that global connectivity, governed by multi-stakeholder institutions and private-sector innovation, serves both American economic interests and American values simultaneously. ICANN, the body that manages the internet’s address system, is the institutional embodiment of this vision: nominally global, structurally American-adjacent, and deeply resistant to intergovernmental control. Washington has treated any challenge to this architecture as, at minimum, an inconvenience, and at most, an act of geopolitical aggression.
China has built an alternative. Its “sovereign internet” model — enforced through the Great Firewall, operationalised through data localisation mandates, and exported through the Digital Silk Road — proceeds from the premise that a state has the same right to control its digital territory as it does its physical borders. This is not merely defensive posture. Beijing has systematically constructed its own parallel ecosystem — WeChat, Baidu, Alipay, Huawei’s infrastructure — capable of sustaining a technologically sophisticated society entirely insulated from Western platforms. It has then offered that model, along with the infrastructure to run it, to dozens of nations in the developing world. Every surveillance system deployed under a Belt and Road digital contract is, in effect, a vote for the Chinese governance model at the ITU.
Europe occupies the most paradoxical position of the three. Brussels does not seek to fragment the internet — it insists, with regularity, on the importance of global openness. Yet the General Data Protection Regulation, the Digital Markets Act, the NIS2 Directive, and the emerging AI Act collectively constitute the most consequential unilateral reshaping of internet governance since the invention of the browser. The EU’s “Brussels Effect” — its capacity to set de facto global standards through the gravitational pull of its market — is producing fragmentation as a side effect of regulation, even as European policymakers denounce fragmentation as a threat. This is not hypocrisy. It is the inevitable consequence of three large actors optimising for incompatible values within a single shared network.
Analyst note
The term “splinternet” obscures more than it reveals. Internet fragmentation is not a single phenomenon but at least three distinct dynamics occurring simultaneously: inward sovereignty (states controlling domestic access), outward projection (states exporting their governance models), and regulatory divergence (liberal democracies producing incompatible compliance regimes). Treating these as one trend produces misleading policy prescriptions.
The Institutional Battlefield No One Is Watching
While the great-power competition over internet governance plays out in headlines about TikTok bans and Huawei exclusions, the more consequential struggle is unfolding in rooms that attract far less attention: the working groups of the International Telecommunication Union, the ICANN policy forums, and the UN’s Internet Governance Forum.
By 2026, the IGF has become what it was always at risk of becoming — a space of dialogue rather than decision. It convenes. It debates. Its outputs remain non-binding. Meanwhile, the real agenda-setting has migrated into security legislation, trade instruments, and bilateral agreements, where speed and enforceability take precedence over consensus. Governments have learned that the fastest way to shape internet governance is not to win arguments at the IGF, but to pass domestic laws with extraterritorial effect and wait for other jurisdictions to respond.
The ITU, long regarded as a venue for technical coordination, has become a diplomatic arena where China has systematically accumulated influence — placing its officials in senior positions, advancing proposals for a state-centric internet protocol architecture, and building coalitions among developing nations who see the existing model as an inheritance of American hegemony rather than a neutral technical arrangement. These nations are not wrong in their diagnosis. They are simply drawing conclusions that Washington finds inconvenient.
The fastest way to shape internet governance is not to win arguments at the IGF. It is to pass domestic laws with extraterritorial effect — and wait for the world to comply.
Infrastructure as Deterrence
What has changed most sharply in the past two years is the strategic framing of internet infrastructure. Undersea cables, internet exchange points, cloud server locations, and domain name systems were once discussed primarily as technical and commercial assets. They are now embedded, formally, in hybrid threat doctrine. EU-NATO joint declarations treat internet infrastructure as a dual-use asset — simultaneously civilian substrate and potential target. The United States has begun classifying certain subsea cable routes as matters of national security. Beijing controls, through state-adjacent enterprises, a significant and growing share of the world’s submarine cable infrastructure.
The logic is straightforward: in a conflict scenario, the ability to degrade an adversary’s connectivity — or to sustain one’s own in the face of attack — is a meaningful military capability. Russia’s 2021 Runet test was not an eccentricity; it was preparation. The question facing every significant state actor is no longer whether to develop this capability, but how quickly to do so and at what cost to the civilian connectivity that their economies depend upon.
This militarisation of internet infrastructure is producing a second-order effect that has received insufficient analytical attention: the compression of the space available for genuinely neutral governance institutions. ICANN, the Regional Internet Registries, and the major standards bodies have always operated on a premise of technical neutrality — the idea that the protocols of the internet are engineering problems, not political ones. That premise is now under sustained and, increasingly, successful assault. An institution that was designed for a world of cooperative technical coordination is poorly equipped for a world in which infrastructure is a lever of deterrence and response.
The Sovereignty Theatre — and What Lies Behind the Curtain
There is a significant gap between the sovereignty that governments claim and the sovereignty they can actually exercise. Encryption standards have advanced faster than state surveillance capabilities. Virtual private networks remain widely accessible in markets where they are nominally banned. Data localisation mandates are routinely circumvented by sophisticated corporate actors who understand, better than most regulators, the technical architecture of the systems they are supposed to control. Much of what presents itself as digital sovereignty is, in practice, sovereignty theatre — laws that exist on paper but are honoured in their breach by the technical reality of the network.
This gap matters, but not in the way optimists typically suggest. The appropriate conclusion is not that the splinternet is overstated. It is that the fragmentation is structural and durable even where it is incomplete. A Great Firewall that leaks around the edges through VPNs is still a Great Firewall. A data localisation regime that sophisticated multinationals circumvent still raises the cost of entry for foreign competitors and concentrates leverage in the hands of the state. Imperfect sovereignty is still sovereignty. And imperfect fragmentation still fragments.
Analyst note
The resilience of the global internet at the user level — the appetite for connection that crosses digital borders — is real and should not be dismissed. But it is also a lagging indicator. The infrastructure-level fragmentation being built today will constrain user-level choices tomorrow. The walls being constructed now are not yet visible to most of the people who will eventually live inside them.
What This Means for Diplomacy, Commerce, and Security
For diplomatic practitioners, the fragmentation of internet governance is not a background condition — it is a primary arena of great-power competition, deserving the same analytical seriousness as nuclear posture or trade policy. The failure to engage with ITU processes, ICANN reform debates, and bilateral digital trade negotiations as geopolitical instruments has left significant ground ceded to actors with clearer doctrines and longer time horizons.
For businesses operating across multiple jurisdictions, the practical consequence is a compliance environment of growing incoherence. A company subject to GDPR, US data surveillance law, and Chinese data localisation requirements is not navigating a single internet — it is navigating three, simultaneously, with contradictory obligations at every layer of its technical stack. The compliance cost is real. The strategic risk — that data held in one jurisdiction becomes a liability in another — is underappreciated.
For security professionals, the fragmentation creates both new vulnerabilities and new ambiguities. A more nationally segmented internet is, in some respects, a more defensible one — smaller attack surfaces, clearer jurisdictional accountability, more legible threat environments. It is also one in which the norms and institutions that have historically constrained state behaviour in cyberspace lose their legitimacy and their reach. The rules that governed the unified internet do not automatically transfer to the splinters.
Bottom line assessment
The splinternet is not a future risk. It is a present condition, advancing across multiple vectors simultaneously. The relevant question is no longer whether the internet will fragment, but how deeply, along which fault lines, and with what consequences for the institutions — diplomatic, commercial, and technical — built to govern a network that no longer exists in the form they were designed to manage. States that treat internet governance as a technical backwater do so at compounding strategic cost. The map of cyberspace is being redrawn. The states with doctrine — however incompatible with one another — will determine its new borders. Those without will find themselves governed by boundaries they had no hand in drawing.
Cyber Diplomacy Internet Governance Digital Sovereignty Geopolitics Splinternet ICANN Vladimir Tsakanyan
Leave a comment