The Digital Strait of Hormuz
Fibre optic networks are heavily concentrated in geography and control
For three decades, the digital economy operated under a utopian illusion: that the internet was a borderless, decentralized commons, immune to the geopolitical friction of the physical world. Because digital trade felt ethereal, we forgot that it requires physical rails.
Today, that utopian dream is colliding with reality. The internet is facing its “Strait of Hormuz” moment, and the primary chokepoints could be the fiber optic cables.
Think of these cables as the shipping lanes of the digital economy, and data centers as their ports. If these lanes are blockaded, severed, or heavily contested, digital trade (~20% of global GDP) is disrupted. The cascading second-order effects - ranging from the balkanization of the internet to a surge in sovereign infrastructure investments - are creating a massive, unrecognized paradigm shift.
The vulnerability lies in an astonishing concentration of both geography and ownership:
95% of global internet traffic runs on roughly 570 undersea cable systems.
Hyperscalers control from 60-90% of capacity depending on the route.
~20% of global data flows through the heavily contested Red Sea corridor linking Asia and Europe.
Supply Chain Duopoly: Just three companies—SubCom (US), ASN (France), and NEC (Japan)—control 87%of cable manufacturing.
Maintenance Fragility: There are fewer than 80 repair ships globally. A geopolitical blockade of territorial waters can turn a routine, weeks-long fix into a catastrophic, months-long outage.
The Topography of Control
The historical precedent for this is the British Empire’s “All Red Line” - a 19th-century telegraph network designed to only touch British soil to prevent wartime interception. Today, we are witnessing the construction of a modern All Red Line, dominated by US tech giants.
US hyperscalers now control 90% of capacity on the trans-Atlantic route. Traditional European operators- who once dominated these waters- hold a rump 2% share. This creates a unilateral risk for Europe: its digital trade, autonomous and digitally connected hardware, and payment rails are completely reliant on infrastructure governed by the US CLOUD Act.
The trans-Pacific route is a little more contested. Hyperscalers hold roughly 60% of capacity, while state-aligned Asian operators hold 25%. Geopolitical risk premia are visibly redrawing the map: cables commissioned since 2020 routinely exclude Chinese carriers, Google controls over a third of the Pacific’s capacity, and routes are intentionally bypassing the South China Sea in favor of Indonesian waters.
Meanwhile, as the Suez-to-Bab al-Mandab corridor proves highly vulnerable (with 25% of traffic disrupted by Red Sea cable cuts in 2024), Africa is quietly emerging as the critical new middle point for global traffic, sparking a proxy war of parallel network build-outs between the US and China.
The White Space for Builders and Capital
When physical shipping lanes degrade, trade stops. When digital infrastructure degrades, traffic re-routes - but at the cost of severe economic friction. As markets and nations factor these vulnerabilities and their increased probability into their strategic calculations, the “splinternet” starts to become a physical reality.
As much as it is a big risk for the current form factor of the Internet and services and trade running on top of it. For the architects of tomorrow, this balkanization potential presents an opportunity both in terms of building the infra as well as the services enablement layer. The mandate is clear: build redundancy.
We are entering a super-cycle of CAPEX investment aimed at circumventing these chokepoints. The immediate beneficiaries will be Edge Computing (keeping data processing onshore), Sovereign Cloud Architectures, Localized Essential Digital Services and Products and Space-based Communications.