The death of the open internet: what replaces the world’s last shared space?
Shutdowns, censorship, and undersea fragility are dismantling the world’s last shared space. The open internet is dying—and what replaces it will shape global power for decades.
For much of its short life, the internet has seemed like the closest humanity had come to a common, global space. Borders have faded, our sense of planetary distance lost meaning, and a single protocol has come to connect the world’s information, conversations, interactions, and markets. Perhaps it hasn’t been a utopia, but it has been universal enough to make geography feel negotiable.
Lately, it feels like the illusion is breaking. The open internet, which was built on interoperability, neutrality, and shared standards, is having a rough time. In its place, a more fragmented system is emerging: national intranets, platform walled gardens, and fragile physical networks governed more by power than by principle. The worldwide web, once imagined as a public commons, is now being carved into zones of influence.
This change is neither abstract nor future tense, since it’s happening in three dimensions at once: political, physical, and architectural. And across the past year, three events spanning Africa, the Middle East, and global infrastructure, all show how fragile the “worldwide” web has actually become.
Shutdown as policy
An internet shutdown is, as it suggests, the disruption of access to online services or networks. This can happen at partial or complete levels, whether by cutting mobile data, blocking key platforms, throttling speeds down to unusable levels, or even blanket disconnecting entire regions. Governments most often justify them as emergency measures to curb unrest or misinformation, but in practice they’ve become instruments of political control. In 2024, governments imposed 296 internet shutdowns across 54 countries, the highest total since records began (Access Now). Twenty-one of those occurred in Africa alone, from Sudan to Senegal, often coinciding with elections or unrest (The Guardian). These aren’t isolated emergencies, it’s very clear they are becoming commonplace, intentional tools of coercion.
Furthermore, since we now live and interact so much online, shutdown erases not just connectivity but civic space. Messaging platforms vanish, banks and hospitals go dark, students lose access to classes, and journalists are left shouting into the empty void. The consequences of this outlast the outages themselves, and as a reaction we see that businesses relocate, trust erodes, and citizens learn that access to information is conditional on the prevailing political mood.
To push the point further, the deeper shift is psychological. Once you understand that your connection can vanish overnight, the internet ceases to be an open utility, a place to be and to get things done, and instead becomes an instrument of permission. The promise of borderless communication has reversed on itself, and the same networks that once dissolved frontiers now rebuild them.
The limits to connectivity
If national or regional shutdowns mark the normalization of selective control for coercive governments, then Iran’s June 2025 blackout revealed the comprehensive, blanket approach. Amid escalating tensions with Israel, authorities simply severed the country’s external connectivity, effectively pulling the plug to isolate tens of millions from the global network and route all traffic through the state-run “National Information Network” (Le Monde). VPNs were blocked, satellite links jammed, and foreign services replaced with state clones.
This was a strategic rehearsal, a demonstration of how a modern state can still detach itself from the world and continue to function domestically. Iran has proven that national-scale disconnection is now a viable governance tool, and we have seen the same pattern appear in Myanmar, Russia, and parts of India. The connection between these examples is that they seek to redefine sovereignty: no longer simply the control of territory and political geography, but the control of networks as an extension of this.
The open internet once promised an unspoken degree of asymmetry, of unfettered openness, so that citizens could access information even when states preferred silence. There was always a window to get in (or out) through the basement if the front door was shut. Now, however, web architecture and policy are aligning to reverse that freedom. A kill switch culture is taking hold.
A very physical infrastructure
Beneath the metaphors of clouds and networks lies something bluntly tangible: the internet is a system of cables, glass, and steel, and that system is beginning to show stress. In September, four major undersea cables running through the Red Sea were severed, reportedly by accident via a dragged anchor from a commercial vessel. The break cut or slowed data flow between Europe, Asia, and Africa for several days, forcing rerouting through longer, slower pathways. It was unintended, but it still exposed a structural truth we rarely want to confront: the digital world rests on a physical skeleton (Associated Press).
Those thousands of cables, laid down in predictable corridors, form the planet’s circulatory system for information. They run along the same maritime chokepoints that define oil routes and trade flows: the Bab el-Mandeb, the South China Sea, the Mediterranean gateways, even the Arctic passages, now thawing into geopolitical theaters. The routes that once moved goods now move bits, and like their older predecessors, as very physical constructs they are equally as vulnerable to pure accident, intentional sabotage, or strategic territorial ambition.
We talk about “the cloud” as if it floats above the real-world reality of politics, but it is a digital construct tethered to geography, to terra firma, as firmly as any port or pipeline. And, as tensions rise in the Red Sea, the South China Sea, and the high north, these fiber conduits are no longer just infrastructure, they are physical leverage. A single mishap or a deliberate strike can fracture the network faster than any act of digital censorship, like debating the terms of digital engagement when, suddenly, the plug is pulled from the wall.
The open internet was always physical; we just nurtured the illusion that it wasn’t.
Related reading: Wired planet: can a global power grid survive national politics?
Fragmentation by design
These fractures are the visible symptoms of a deeper, intentional segmentation of the network itself, to re-engineer it through the laws, markets, and power structures that govern it. Over the past decade, and especially since 2020, digital policy has notably shifted from connecting the world to protecting it. The European Union’s Digital Services and AI Acts, the United States’ export-control and data-security regimes, and China’s insistence on data localization all serve different political agendas, yet they converge toward a similar outcome, namely, the partitioning of information flow along national and ideological lines.
Related: The battle to govern AI: How Europe regulates, America resists, and China advances
Each framework makes sense on its own—protecting privacy, security, or sovereignty—but together they’ve carved the web into semi-connected zones where access depends on jurisdiction. The “splinternet,” once an academic warning, is now real. Freedom House’s Freedom on the Net 2025 report says it plainly: the internet has shifted from open participation to pre-approved permission. (Freedom House).
This isn’t all at the governmental level, of course. While we see the redrawing of digital borders through law, corporations are doing the same through code. The open protocols that once allowed anyone to publish and link freely have been superseded by a tech-led culture of pure capitalist drive, with platform ecosystems being designed to monetize every gesture of connection. Search is no longer the simple quest to locate reliable information, rather, it now leads not outward but inward, into proprietary indexes, digital siloes of control. Social networks go further, grabbing our attention to collapse the wider web into algorithmic corridors that narrow our experience. AI interfaces take this even further, having become many people’s first stop for information, yet this still equates to the filtering of knowledge through opaque licensing agreements and moderation policies, none of which is uniformly governed in any meaningful sense at the intergovernmental level.
We did have a web that, especially prior to Google’s search monopoly and the onset of social platform algorithms, constituted a loose connective tissue of hyperlinks and open discovery. This has, however, been largely replaced by feeds, prompts, and commercial mediation, little of which the user is ever aware. This enclosure is also self-reinforcing, since the more data these systems capture, the more indispensable they become, and the less incentive users have to look beyond their walls. Sure, there is still some rhetoric of openness, but it is increasingly marginalized to the corners of the web, floating above and around privately owned architectures whose incentives point elsewhere, firmly toward greater control and profit maximization.
Freedom of expression remains, if we pursue it, but increasingly within fenced gardens where the landlords own the rules.
Architecture of control
At a technical level, the open internet does still exist. The underlying protocols—TCP/IP, DNS, BGP—remain universal, and packets still cross borders. However, the governance wrapped around them has changed, and with it the intent. A recent Sciences Po study describes this shift as the “outward turn” of fragmentation, meaning that instead of merely restricting domestic content, states now project and exert digital power abroad by funding the cables, satellites, and data centers that embed their influence into the global network itself. Connectivity has become a tool of diplomacy and leverage, and what began as neutral infrastructure is now put to use as foreign policy, turning the network from a public utility into a contested strategic terrain.
The original spirit of the web, of the early internet, was trust. It was built on the idea that openness would be the driver of innovation and transparency would self-reinforce to keep abuse in check. Now, the firewalls, filters, and algorithmic safeguards against misinformation have become the default settings, turning the web into a place for conflict and control. Universal access? This is being choked under the layers of verification, compliance, and surveillance. What used to be a free, open sandpit where you could play and even make up the rules, it now feels like you have to ask the local bully first for permission to be there.

The world after openness
So, to our title question. What replaces the open internet? Well, it will not be a single successor system but a mosaic of incompatible ones, based on differing or competing interests. Authoritarian systems like Russia’s Runet or Iran’s National Information Network no longer just censor; they substitute, building sealed versions of the web that look functional from the inside while totally cutting off the outside view. Corporate ecosystems, however, are no less ruthless in their approach, and whether it be Apple, Meta, or Tencent, they do something similar through commerce, enclosing users inside profit-driven environments, where what you get to see and how people get to see you is now a commercial decision wrapped in the illusion of being, somehow, ‘social’.
Meanwhile, regional regulatory models such as the EU’s digital-sovereignty rules or India’s localization mandates redraw borders around data, storage, and access; and at the fringes, civic technologists keep trying to find weak points, testing decentralized networks that route around both government and corporate chokepoints. So, what’s the result? It feels less a single internet than a patchwork of managed domains, with corporate or governmental authority stacked at the top, fragmentation seeping upward from the bottom, and what once felt like a commons squeezed into whichever gaps it can find.
The implications of this are significant. Losing a shared digital space means losing one of the few remaining arenas of unfiltered collaboration, communication, of expression. The open internet once allowed scientists, journalists, and ordinary citizens to exchange information faster than politics could hope to interfere, and its slow disintegration reverses that significant advantage. The version we are seeing now is a less challenging, bubbled space: discussion narrows to familiar languages or regional/ethnic loyalties; misinformation cycles faster within walled echo chambers; and inequality grows as rich regions build backup networks while poorer ones depend on borrowed or filtered access.
A mirror of global hierarchy? Are we simply reconforming the web to emulate our emerging multipolarity, our fractures? We’ve definitely been here before, when other new communications technologies conformed to the needs of the powers wielding them: postal unions that fractured into colonial circuits, the telegraph lines nationalized into empires. It seems that the internet’s next chapter could read the same: a shared invention broken into private domains.
Designing what comes next
Perhaps the feeling of increasing digital enclosure, of being fenced in, is in keeping with the very shifts in politics we are seeing, ironically to a large extent via the net. As with the challenges facing open democracy, the decline of the open internet is in no way inevitable, but we must acknowledge that arresting it means putting up a firm resistance to the political and commercial forces that drive consolidation. Real openness depends on the less glitzy qualities rarely rewarded in the short term: redundancy, transparency, and shared governance. There are efforts, it should be noted, to restore them.
Related: The return of the mega-factory: how manufacturing power is concentrating again
Decentralized web projects, community mesh networks, and public-interest digital infrastructure aim to re-distribute control to the edges rather than the core. There have been calls in recent years for “re-decentralization” shows that technical renewal remains possible, though political will is scarce (HTR). What comes next depends on what kind of governance, and restraint, we choose.
As a final note, this pattern is familiar, and something we keep arriving at: censorship, monopolies, and regulatory walls create control first and fragility later. It’s a shortcut to the tougher work that must be undertaken, where the wonderful complexity of diversity and challenge is embraced, not the bubbled, sterile, narrow corridor to stifled voices and monetized solutions. Only cooperative, transparent systems endure; the rest will ultimately collapse under their own design.
Read this. Notice that. Do something.
Read this: Freedom on the Net 2025 – Freedom House documenting the 15th consecutive year of global internet-freedom decline.
Notice that: Access Now / #KeepItOn Report 2024 on how shutdowns have become routine tools of political control.
Do something: Read Internet Society – The Internet Needs a New Commons outlining governance reforms for re-decentralization and public accountability.
Previously on GYST: The return of the mega-factory: how manufacturing power is concentrating again
Next up: Tariffs, trade & tension: supply-chain shock #4.0