When the Anchor Tenant Builds Its Own Ship: Africa’s Subsea Cable Reckoning

By /

Here is a fact that still surprises people. In 2009, the East African coast — from Sudan and Kenya down through Tanzania and Mozambique, hundreds of millions of people — had not a single submarine cable of its own. Everything moved by satellite: slow, scarce and ruinously expensive. Then, in July 2009, a ship finished laying SEACOM down the Indian Ocean coast, and a region that had been paying satellite prices for a trickle of bandwidth suddenly had a motorway.

Fifteen years later, the pendulum has swung so far the other way that the problem is no longer scarcity. It is what to do with abundance — and, for the companies that built the first motorways, what happens when the biggest customers stop paying tolls and build their own roads.

This is a story about glass, ocean and power. But underneath it is a sharp commercial question every operator, investor and government touching African connectivity should be asking: as Equiano, 2Africa, Medusa and a wave of new systems land, what is the older generation of cables actually worth — and who still needs them?

Three waves, three different worlds

Africa’s subsea story runs in three distinct eras, and they could not be more different.

The monopoly era. For most of two decades, West Africa’s connection to the world ran through essentially one cable: SAT-3/WASC, live from around 2001–2002, carrying a modest ~120 Gbit/s and controlled by a tight club of incumbent telcos. Its predecessor, SAT-2, had served since 1993. With no alternative, landing-station owners could — and did — charge extraordinary margins to move traffic between the sea and the land. East Africa didn’t even have that; it had satellite.

The liberation wave. Then, in a remarkable four-year burst from 2009 to 2012, the map filled in. SEACOM (2009) and TEAMS opened the East African coast; EASSy (2010) added redundancy; MainOne (2010) and GLO-1 lit up West Africa; and WACS (2012) and ACE (2012) brought serious capacity down the Atlantic seaboard. Wholesale prices collapsed, the SAT-3 monopoly broke, and Africa’s internet economy was, finally, buildable. The firms that led that wave — SEACOM, WIOCC, MainOne, Angola Cables and the consortium members — became the backbone wholesalers of the continent.

The hyperscaler era. And now the third wave, which is not really about connecting countries at all. Google’s Equiano landed in 2022 with roughly twenty times the capacity of any previous cable serving the region. 2Africa completed its core route in November 2025 and, with the “Pearls” extension due in 2026, will run to ~45,000 km across 46 landing points and 33 countries, at up to 180 Tbit/s — the longest open-access subsea system in the world. For scale: the planet’s circumference is 40,075 km. 2Africa is longer than the world is round, and it wraps Africa almost entirely.

Three eras: scarcity, liberation, and now abundance. That last shift — and who owns the abundance — is the whole story.

The new giants, and the three ways to own an ocean

Here is where the common shorthand (“the hyperscalers are taking over”) needs unpacking, because the ownership picture is more interesting than that. Three distinct models are emerging at once.

The wholly-owned hyperscaler cable. Google owns Equiano outright, and its new Umoja system — running overland from Kenya through Uganda, Rwanda, the DRC, Zambia and Zimbabwe to South Africa, then subsea to Australia — is the first direct Africa–Australia route. Its overland leg (built with Liquid) is complete, but the subsea crossing to Perth is still under construction, with completion now reported to have slipped toward 2027. Meta’s Project Waterworth, unveiled in early 2025, will be the world’s longest subsea cable at ~50,000 km: a 24-fibre-pair system spanning five continents including South Africa, diving to 7,000 m. These are private roads, built and owned by one company for its own traffic.

The hyperscaler-anchored consortium. This is the crucial nuance, and it is exactly right to flag it: 2Africa is not a Meta-owned cable. It is a consortium. Meta leads and anchors it, but the members include Bayobab (MTN), Orange, Vodafone, center3 (stc), China Mobile International, Telecom Egypt and WIOCC. So African and international operators sit on the ownership table alongside the hyperscaler — sharing the cost, the capacity and the risk. This hybrid is arguably the more durable model for Africa, because it keeps operators in the room rather than reducing them to customers.

The independent open cable. And then the newcomers that are neither telco-club nor hyperscaler. Medusa — owned by Spain’s AFR-IX Telecom, an 8,700 km, 24-fibre-pair system rated at up to 480 Tbit/s — began as a Mediterranean cable and is now extending down the Atlantic coast into West Africa, backed by an EU Connecting Europe Facility grant and US Trade and Development Agency funding, with landings planned via Gabon (Port-Gentil) and the DRC. Medusa matters because it is an open-access, publicly-backed alternative to both the telco consortia and the hyperscalers — a different funding logic entirely.

The common thread across all three: the new cables are bigger, more direct and more resilient by design. But they are being built by a wider, stranger cast than the old telco consortia — and overwhelmingly anchored by the companies that generate the traffic, not the ones that used to sell it.

The fuller map: three coasts, and a bridge to the Americas

Step back from the headline systems and the continent’s real subsea geography comes into focus — three distinct theatres, each with its own logic.

The Atlantic seaboard is the busiest and oldest: SAT-3, then WACS, ACE, MainOne and GLO-1, now overlaid by Equiano, the western arc of 2Africa, and Medusa’s coming extension to Gabon and the DRC. This is where capacity is most abundant and wholesale margins are most compressed.

The Indian Ocean seaboard and its islands tell a different story. Alongside the mainland systems — SEACOM, EASSy, TEAMS, DARE1 and PEACE — sits a dense mesh of regional cables knitting the island economies together: LION and LION2 (linking Madagascar, Réunion and Mauritius, with LION2 extending on to Mayotte and Kenya), the veteran SAFE, METISS (linking Mauritius, Réunion and Madagascar down to South Africa), MARS (Mauritius to Rodrigues), and the newly landed T3, a ~3,200 km Mauritius–South Africa system that is effectively a revival of the shelved IOX project. Mauritius has used this density deliberately, positioning itself as a low-latency data and financial-services hub — connectivity as economic strategy.

And then the bridge west. The least-known and most strategically intriguing theatre is the South Atlantic, where Africa has built the first direct cables to South America. SACS, owned by Angola Cables, runs 6,165 km from Sangano in Angola to Fortaleza in Brazil — the first commercially operational submarine cable across the South Atlantic — and, stitched to the onward Monet system, gives Angola a route to the Americas that never touches Europe. SAIL, owned by Camtel and China Unicom (and built by Huawei Marine), links Kribi in Cameroon to the same Brazilian hub at almost exactly the same time. These cables are not about capacity abundance; they are about latency and routing independence — reaching the Americas directly, and no longer sending every African byte through a handful of European chokepoints. That is a sovereignty argument dressed as an engineering one — and it is exactly the logic now driving the newest global builds.

Why the giants went to sea

The shift that reframes everything: subsea cables used to be built almost exclusively by consortia — dozens of telcos pooling money over four-to-six-year negotiations, each taking a slice of capacity to resell. It was slow, political, and it made the cable itself the business.

The hyperscalers broke that model. When your own traffic outstrips what any carrier needs, and bandwidth demand doubles roughly every 18 months, buying wholesale capacity becomes both expensive and strategically limiting. So Google, Meta, Microsoft and Amazon started building their own — cutting procurement from six years to under three, and controlling every route and landing point themselves.

The scale of the takeover is staggering. Over the past decade, the four hyperscalers’ share of international subsea capacity has risen from around 10% to roughly 70%. They now own or fund 30-plus cable systems and carry the majority of the world’s transoceanic traffic on their own infrastructure. The cable is no longer a product they buy. It is plumbing they own — increasingly, plumbing built to connect their data centres and feed their AI.

The world has done this before

Before assuming Africa’s abundance is a permanent gift, it is worth remembering that the world has already lived through exactly this movie — and it did not end kindly for the incumbents.

Consider TAT-8, the first transatlantic fibre-optic cable, laid in 1988 by a consortium of AT&T, BT and France Telecom for about $335 million. It was designed to last 25 years. Its capacity was saturated within 18 months as the early web exploded. It was retired in 2002 after just 14 years, when a fault made repair uneconomic — and right now, in 2026, it is being physically hauled off the seabed and recycled after nearly four decades underwater. The cable that made the global internet possible is being ripped up.

Then came the cautionary tale proper. In the late 1990s, capital flooded into subsea cables on the assumption that demand would rise forever. It over-built spectacularly. When the dot-com bubble burst, investment in submarine cable projects fell roughly twelve-fold between 2001 and 2003, new deployments fell ten-fold, and a wave of the era’s champions — Global Crossing, WorldCom, Viatel — went bankrupt in the “bandwidth glut.” Supply had raced miles ahead of demand, and the cable-laying industry collapsed almost overnight.

The legacy is still being cleared. Of an estimated two million kilometres of decommissioned subsea cable worldwide, most has never been recovered — and close to a million kilometres more is set to be retired by 2040, the bulk of it within the next decade, as high-fibre-count systems make the older, power-hungry cables uneconomic to maintain.

The lesson for Africa is not that abundance is bad. It is that abundance is not the same as value — and that the assets which get stranded first are the mid-life cables whose economics depended on being scarce.

The reckoning for Africa’s incumbents

Put yourself in the position of a cable built in the liberation wave — a SAT-3, a WACS, an ACE, a first-generation SEACOM. For fifteen years your best customers were the content providers and the international carriers buying wholesale capacity to reach African users. They were your anchor tenants. They paid the bills.

Now those anchor tenants have built, or bought into, their own ships. Google reaches Africa on Equiano and Umoja. Meta anchors 2Africa and is building Waterworth. They no longer need to lease your capacity — and when the largest, highest-volume buyers exit the wholesale market, volumes fall and prices collapse at the same time, because the remaining capacity is suddenly abundant.

Layer on the physical clock. Submarine cables are built to a design life of roughly 25 years — though, as TAT-8 showed, many are retired well before that. SAT-3 (2001–2002) is now brushing that ceiling; WACS and ACE (2012) are middle-aged. But as TAT-8 showed, a cable does not need to physically fail to become commercially stranded — it simply stops earning enough to justify the maintenance, the marine spares and the repair-ship standby fees. Commercial end-of-life now arrives well before technical end-of-life.

So the honest answer to “will the older cables keep their hyperscaler and OTT anchors?” is: mostly, no. The hyperscalers are becoming competitors, not customers. The pure-play capacity wholesaler — the model that defined the liberation wave — is the most exposed asset class in African connectivity, and the part of the market where we would be most careful about valuations built on yesterday’s wholesale revenue.

Repositioning: everyone is moving up the stack

The good news is that the smart incumbents saw this coming, and the response is instructive. It is the same move, again and again: stop selling raw bandwidth, and start selling everything around it.

WIOCC is the clearest template. Once a pure subsea-capacity wholesaler, it raised $200 million and spun up Open Access Data Centres (OADC) — buying capacity across Equiano, EASSy and 2Africa while building carrier-neutral data centres in Lagos, Durban and beyond. It stopped competing with the hyperscalers and became the landing-and-interconnection partner they need on the ground.

MainOne took the same logic to its conclusion: it sold to Equinix for $320 million in 2022 and became a data-centre and interconnection platform; Equinix has since committed up to a further $390 million to African expansion. The cable is now a feeder to the colocation business, not the business itself. SEACOM pivoted from wholesale bandwidth toward enterprise and managed services; Angola Cables leaned on its South Atlantic routes and its AngoNAP data centres.

The pattern is unmistakable: the cable becomes the loss-leader; the margin migrates to the landing station, the interconnection point, the data centre and the enterprise service sitting on top of it. Value is draining out of the wet plant and pooling at the edges — where the cable meets the land, the cloud and the customer. The winners are repositioning as the neutral, open-access glue between global cables and African demand. The losers are still trying to sell megabits.

Why this is, in the end, a resilience story

For all the talk of abundance, the single most important reason Africa still needs many cables — old and new — has nothing to do with price. It is resilience.

The continent has already had its wake-up call. In March 2024, a suspected undersea landslide in the Trou Sans Fond canyon off Côte d’Ivoire severed four cables at once — WACS, MainOne, SAT-3 and ACE — and knocked as many as thirteen countries offline, weeks after three separate cables were cut in the Red Sea. Millions lost connectivity not because there was no capacity, but because there was no diversity: too much traffic depending on too few paths, converging on the same chokepoints.

This is the paradox at the heart of the story. Hyperscaler abundance is collapsing the commercial value of individual cables — yet the case for having more and more diverse cables has never been stronger. Every additional independent route, every landing that avoids the Red Sea and the Egyptian land-crossing, every terrestrial detour like Umoja’s overland leg, is a resilience asset even if its wholesale economics are thin. And resilience has its own hard constraints: the world’s fleet of cable-repair ships is small and ageing, and a fault off a remote coast can take weeks to fix because the nearest vessel is an ocean away.

So the enduring rationale for subsea investment in Africa is quietly shifting — from capacity to continuity. A cable’s value increasingly lies not in the terabits it can sell, but in the outage it can prevent. That is a very different way to underwrite an asset, and most business cases have not caught up to it.

What it means, and where Digitalthings fits

For investors, a subsea asset’s worth now depends less on its route and rated capacity, and more on who its customers will still be in five years — and on its role in a resilient system. A cable whose anchor tenants are building their own systems is a very different credit from one embedded in a data-centre and interconnection platform, or one that provides a genuinely diverse path. Underwrite the customer base, the up-stack position and the resilience value — not the fibre.

For operators and cable owners, the question is no longer “how much capacity can we sell?” but “where on this value chain do we stand when bandwidth is effectively free?” — landing stations, open-access interconnection, data centres, enterprise services, and the redundancy others will pay to rely on.

For governments, abundance is a gift and a trap: cheaper connectivity, but growing dependence on foreign-owned infrastructure and a concentration of failure points that a single bad day at sea can expose.

This is precisely the terrain we work. At Digitalthings we do the commercial and technical due diligence on subsea, landing-station and data-centre assets — including the decisive, unglamorous questions of remaining commercial life, anchor-tenant risk, capacity-price trajectory and route resilience. Our subsea track record includes West African landing-feasibility studies on Google’s Equiano for investors, and commercial and technical due diligence on the MainOne system for a prospective investor. We advise operators repositioning from wholesale capacity toward interconnection and data centres, and we help investors tell the difference between a cable asset with a future and one quietly approaching its commercial end.

The ocean floor is more crowded than it has ever been. But in African connectivity, as everywhere else, the money has stopped following the cable — and started following the landing.

Prepared by Digitalthings. Figures are drawn from public sources and verified against their publications; cable capacities and dates reflect design specifications and reported launch years. Data is current as of mid-2026.

Sources

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top