The Submarine Cable Question

Cable ownership is data jurisdiction by another name.

The Proof It Can Be Done

EllaLink is an optical submarine cable linking Portugal and Brazil. It is the first direct data route between Europe and South America since the 1990s. Before EllaLink, eighty to eighty-five per cent of data traffic between the EU and Latin America transited North America. Every packet between Lisbon and São Paulo passed through US jurisdiction.

Brazil’s President Dilma Rousseff said in 2014, in the wake of the Snowden revelations, that the cable would be central to guaranteeing the neutrality of the internet.

The model matters. EllaLink was built by Alcatel Submarine Networks, a European manufacturer. It was co-funded by the European Commission. It delivers competitive performance. It is privately funded, carrier-neutral, and open access.

It is shared infrastructure, sovereignty by design, that works and is expanding, with branches to Cape Verde, Mauritania, Morocco, Senegal, and French Guiana in progress.

EllaLink got something right that almost no other European cable project has. The cable lands at Sines, in a facility operated by a Portuguese state-owned company. Sovereign cable, sovereign landing station. But even EllaLink’s data must pass through Equinix data centres in Lisbon and Madrid for interconnection with other networks.

Equinix is incorporated in the United States, subject to the CLOUD Act. The NSA’s UPSTREAM programme collected data at precisely these kinds of exchange points. EllaLink built sovereignty into the ocean crossing and the landing. The gap is one hop further inland, at the interconnection layer. For almost every other European cable, the gap starts at the shore itself.

One cable to Brazil is a proof of concept. The question is why Europe proved it could build sovereign cable infrastructure five years ago, got the model largely right, and then never replicated it.

Europe’s Factory

The answer is not capability. Europe has the world’s leading submarine cable manufacturer. In November 2024, the French government completed the purchase of eighty per cent of Alcatel Submarine Networks from Nokia for €350 million. The French Economy Minister explicitly called it a matter of sovereignty. ASN is not a minor player.

And ASN is not alone. Prysmian Group, headquartered in Milan, is the world’s largest cable manufacturer overall. The Italian government acquired Sparkle, Telecom Italia’s submarine cable operations unit, for €700 million. Of the four companies that manufacture and install ninety-eight per cent of the world’s undersea cables (ASN, Prysmian, SubCom, and NEC), two are European.

Europe controls half the global supply chain for submarine cable infrastructure.

France, Italy, and Spain are consolidating Europe’s cable manufacturing and operations capacity. The supply side is world-leading. ASN builds cables for everyone: Asia-Pacific contracts, Latin American routes, the Mediterranean. Meta’s 2Africa Pearls extension was contracted to ASN before geopolitics forced a halt. ASN builds for whoever commissions it.

Meta commissioned. Europe, for its own sovereign backbone, did not.

France understood the supply side enough to nationalise the manufacturer. Italy understood it enough to acquire the operator. But nationalising the supplier does not help if you never place the order. And placing the order is what Europe has not done.

Who Owns the Wires

US hyperscalers now own approximately ninety per cent of capacity on the transatlantic route and over seventy per cent of international bandwidth globally. Amazon, Google, Meta, and Microsoft now own approximately half of all undersea bandwidth worldwide.

These are not shared infrastructure investments. Google is the world’s largest owner and investor in submarine cable networks, with wholly owned private cables. Meta announced Project Waterworth, a 50,000-kilometre cable that will be the longest in the world, owned by a single company, deliberately bypassing Europe. These cables are built for internal consumption.

They are vertical integration of the physical layer. The companies that dominate European cloud computing now also own the wires that carry the data to those clouds.

The shift happened in two phases. First, European telecommunications operators stopped investing in submarine cables, constrained by financial pressures. Second, hyperscalers stepped in. Their connectivity needs reached a scale where owning a cable was cheaper than leasing capacity from carriers.

The economics were simple: a $300 million cable is trivial against $650 billion in annual capital expenditure.

The consequence for Europe is structural. The thirty-three newest cables provide seventy-four per cent of total capacity landing in EU member states. But these newest cables are overwhelmingly hyperscaler-owned private infrastructure. Legacy consortium cables carry Europe’s public telecoms traffic.

When those legacy cables reach end of life, European operators will need to secure capacity on cables owned by American companies. Europe’s public internet backbone is being replaced by private American infrastructure, and nobody commissioned an alternative.

The European wholesale telecommunications model is collapsing under this pressure. Hyperscalers moved from buying bandwidth to owning cables, making it impossible for European carriers and smaller providers to achieve returns. They did not just compete with European telecoms. They pulled the floor out from under them.

How the Other Side Built It

The hyperscaler advantage was built on guaranteed government contracts. In 2013, Amazon Web Services signed a $600 million cloud contract with the CIA when its estimated yearly revenues were approximately $800 million. Pentagon, NSA, and intelligence community contracts followed at tens of billions in scale, funding infrastructure that has since generated over $100 billion in annual revenue.

American industrial policy has protected domestic digital infrastructure through procurement legislation since the Buy American Act of 1933. Europe has no equivalent.

In February 2026, the US State Department under Secretary of State Marco Rubio issued a classified cable to diplomatic missions across Europe instructing diplomats to counter European digital sovereignty initiatives as a threat to American strategic interests.

The cable identifies submarine cable ownership and digital infrastructure independence as a focus point for diplomatic counter-pressure. Europe is not failing to build sovereign cable infrastructure by accident or inattention. It is failing against active diplomatic resistance from the government whose companies dominate the market. The Active Hand is not a historical episode. It operates now.

Seven Countries Already Moving

This is not, however, a story of complete inaction. Seven European countries are already building pieces of cable sovereignty. None of them coordinated with the others. None is funded at the scale of the challenge. But the pattern they form is more significant than any single initiative.

The pattern has three dimensions. The first is manufacturing and ownership. France nationalised ASN and operates, through Orange Marine, one of the largest repair fleets in the world. Italy has Prysmian and acquired Sparkle for €700 million, building vertically integrated cable capability from manufacturing through operations. Between them, these two countries control half the global supply chain.

The second dimension is route-building. Denmark committed $468 million to a trans-Kingdom cable connecting to Greenland and the Faroe Islands as part of its Arctic defence strategy. Norway is funding the Arctic Way cable to Svalbard as a sovereign defence initiative, and hosts ASN’s SMART cable research team in Trondheim.

Finland is leading the Far North Fiber project through Cinia, an Arctic cable from Europe to Japan via the Northwest Passage. Portugal hosts EllaLink and is deploying the Atlantic CAM project, Europe’s first SMART cable with integrated climate monitoring sensors.

Spain’s AFR-IX Telecom is leading the Medusa Mediterranean cable, a partnership combining ASN (French manufacturer), Elettra TLC (Italian specialist in submarine cable installation and maintenance), and Orange as anchor tenant.

The third dimension is what it proves. Seven countries, acting independently, have each concluded that sovereign cable infrastructure is worth building. They are spending billions without coordination, without a common framework, and without legislated demand to underwrite the investment. The question for European policy is not whether these countries see the value. They already do.

It is whether legislation can create the demand signal that makes every project in the pipeline independently viable, and every future project investable.

Figure 1: Global submarine cable ownership. Hyperscaler-owned cables (red), legacy consortium cables approaching end-of-life (amber), European sovereign cables (blue). Legacy cables carry European public traffic. Their replacements are overwhelmingly American-owned.

The Kattegat Problem

The Kattegat and the Danish Straits are the only maritime access points to the Baltic Sea. Every cable entering or leaving the Baltic passes through Danish waters. Denmark is the gatekeeper of Baltic connectivity.

Danish cables have also served as surveillance infrastructure: in 2021 it was revealed that Danish military intelligence facilitated NSA surveillance of European leaders by providing access to cables in Danish waters.

And yet Denmark has also led on cable protection, passing legislation establishing safety zones in its exclusive economic zone, committing $468 million to sovereign Arctic cable infrastructure, and leading the response when the Yi Peng 3 was held in Danish waters after the Baltic cable cuts. Denmark is simultaneously victim, accomplice, and pioneer.

If the cable sovereignty argument has a geography, it runs through the Kattegat.

When Cables Break

The Baltic Sea has become a testing ground for what cable vulnerability looks like in practice.

In November 2024, two cables were cut almost simultaneously. The BCS East-West Interlink between Sweden and Lithuania, and C-Lion1 between Finland and Germany. C-Lion1 was the only cable connecting Finland to the European continent. On Christmas Day 2024, the Russian tanker Eagle S severed the Estlink 2 power cable and multiple data cables between Finland and Estonia.

These incidents exposed more than physical vulnerability. They exposed a coordination gap. The lack of a NATO-wide protocol meant Finland and Estonia activated national responses, but information was not systematically shared across the alliance. It took hours to identify the suspect vessel.

The gap between the threat and the response is not limited to the Baltic. Globally, the cable repair fleet is ageing out. The EU allocated €20 million for repair capacity. That is less than one per cent of the global need. Average repair times are rising. And in wartime, the commercial repair agreements that the industry depends on become unenforceable.

The United States understood this early. It created the Cable Security Fleet in 2019, subsidising two repair ships with a sovereign access mandate. The programme cost $10 million per year. It guaranteed that in a crisis, American cables would be repaired first. Then it was defunded. Even the United States could not sustain consistent commitment to cable repair sovereignty.

Europe has not even started.

Sovereignty you cannot repair is sovereignty you do not have.

What Flows Through the Cables

The physical vulnerability of submarine cables is not abstract. It maps directly onto financial infrastructure.

SWIFT processes approximately 44.8 million daily messages among over 11,000 institutions, dependent on submarine cables for ninety-eight per cent of its functionality. CLS Bank handles $6.6 trillion in daily foreign exchange settlements. Euroclear manages €35.2 trillion in assets with a recovery time objective of two hours. DTCC processes $2.3 quadrillion in securities annually.

In January 2023, a Nordic exchange disruption lasting 2.7 milliseconds impacted eighty-nine per cent of trading algorithms, caused €2.8 billion in market value loss, and took forty-seven minutes to recover. That was 2.7 milliseconds. A multi-cable outage lasting days or weeks is a different order of magnitude.

London and New York financial centres carry cumulative daily transaction volumes worth $10 trillion through undersea cables, yet few regulators subject such scenarios to formal stress testing.

European dependency in payments, stablecoins, and financial settlement infrastructure rests on the physical layer documented here. Financial sovereignty that transits non-sovereign cables is financial sovereignty on loan.

The digital euro, currently in development, will inherit the same constraint. A sovereign European currency that transits non-sovereign cables is sovereign in name only.

The World Is Proving This Right Now

In March 2026, the US-Iran conflict closed both the Strait of Hormuz and the Red Sea simultaneously for the first time in history. Approximately seventeen submarine cables pass through the Red Sea, carrying the vast majority of data between Europe, Asia, and Africa. Cable repair ships already deployed to fix earlier cuts were forced to suspend operations indefinitely.

The Hormuz crisis is proving every structural argument made here, in real time. Cables concentrated through a single chokepoint. Repair capacity unable to operate in a conflict zone. Financial infrastructure dependent on physical routes that can be closed by events Europe does not control. The vulnerability is not theoretical. It is on the news.

The crisis also demonstrates the alternative. Arctic cable routes pass through waters where NATO allies have presence, where there is almost no commercial shipping traffic, and where insurance premiums are lower because they are not on the Lloyd’s Joint War Committee exclusion list. The Red Sea carries ninety per cent of Europe-Asia data traffic.

A conflict that was foreseeable, in a chokepoint that was known, is now severing connections that had no redundancy. The Arctic routes being planned by Finland, Norway, and Denmark would bypass every one of these chokepoints. The question is why they were not built before the crisis proved they were needed.

The pattern extends beyond the Middle East. Chinese researchers have designed devices explicitly intended to sever submarine cables. The Baltic and Taiwan cable-cutting patterns mirror each other: anchor-dragging as grey-zone warfare. Cable vulnerability is a global pattern, and Europe is among the least prepared of the targets.

Where Sovereignty Ends

Even where Europe builds sovereign cables, a critical gap remains at the shore.

Cable owners are increasingly terminating cables directly inside data centres close to the coast rather than in traditional beach landing stations. The company that dominates European cable landing infrastructure is Equinix, a US-headquartered, NASDAQ-listed company operating approximately 38 subsea-enabled data centres worldwide.

Equinix operates the landing stations or primary interconnection points for Havfrue, the MAREA cable in Spain, and dozens of other cables that serve European connectivity. EllaLink is a rare exception: it lands at Sines in a Portuguese state-owned facility. For almost every other major cable serving Europe, the first point of contact on land is US-owned.

A Microsoft executive admitted under oath to the French Senate that Microsoft cannot guarantee data sovereignty to European customers because of the CLOUD Act. If the provider of Europe’s most widely used cloud services cannot guarantee sovereignty in its own platform, a data centre operator subject to the same jurisdiction certainly cannot guarantee it at the landing station.

The EU’s own Cloud III procurement may qualify AWS Brandenburg as sovereign under a framework that does not assess the full stack.

The European data centre landscape is consolidating under US ownership. Interxion, one of Europe’s largest operators, was acquired by Digital Realty for $8.4 billion. European alternatives exist: DE-CIX in Germany, Bulk Infrastructure in Norway, Orange in France, Sparkle in Italy. But the default industry practice routes cables into US-owned facilities because they have scale.

Without a mandate to use European landing stations, the default will persist.

This introduces a principle that extends beyond cables to the entire series. A stack’s sovereignty position is inherited from its least sovereign component. If the cable is Position 1 but the landing station is Position 4, the effective sovereignty of the data path is Position 4. You cannot buy sovereignty one layer at a time.

The full data path for a typical European cable, such as Havfrue running from Denmark to the United States, runs entirely through American-controlled layers.

Every layer is Position 4. The cable is American-owned. The cable lands at a neutral facility, but traffic routes through American Points of Presence (Equinix or similar providers) before reaching its destination. The cloud above it is American-operated. The effective sovereignty of the entire data path is determined by its least sovereign component.

EllaLink proved it is possible to get the first two layers right: European cable, European landing station. But EllaLink is one cable. Havfrue is the norm. Any legislation that aims to guarantee sovereign infrastructure must define sovereignty as a property of the complete stack, not of any individual component.

Otherwise, the same pattern that allowed GDPR to be met by minimum viable compliance will repeat itself at the infrastructure layer.

What Europe Should Build

The preceding sections establish three facts. Europe has the manufacturing capability to build sovereign cables. It has not used that capability for its own backbone infrastructure. And every day of inaction deepens a dependency that extends from the ocean floor to financial markets.

This section proposes a response. The series has spent fourteen papers diagnosing the structural condition. The cable question is where diagnosis becomes architecture, because this is the layer where Europe already has every component it needs.

The first fix is on land. EllaLink got this right. It lands at Sines, in a Portuguese state-owned facility. Sovereign cable, sovereign landing station. But EllaLink is the exception. Most European cables land in Equinix or other US-owned data centres, re-entering American jurisdiction at the shore.

This is fixable now, without new legislation, without a decade-long programme, without consensus among twenty-seven governments. The landing station is a data centre. European-owned data centres exist. The requirement is simple: sovereign cables must land in sovereign facilities.

Italy demonstrated the model when it acquired Sparkle for €700 million in February 2025, bringing cable landing and operations under sovereign control. France has sovereign facilities through Orange. Bulk Infrastructure operates in Norway. DE-CIX operates in Germany. The capacity exists. What is missing is a requirement: sovereign cables must land in sovereign facilities.

Where a European-owned landing station does not exist at a critical termination point, one should be acquired or built. Where necessary, nationalised. The cost of acquiring a landing station is trivial compared to the cost of the cable it serves.

Allowing a €80 million sovereign cable to be compromised by a landing station that could be secured for a fraction of that amount is not a policy failure. It is an oversight that should be corrected before the next cable is commissioned.

Europe nationalised the cable manufacturer but most traffic still routes through American Points of Presence before reaching its destination, winning one layer of the stack while losing the next.

Investment, not expenditure. A sovereign European cable programme covering transatlantic, Arctic, Mediterranean, and intra-European backbone routes, plus sovereign landing stations and a repair fleet, is the asset Europe has not built. Cable infrastructure is an asset. The money builds capacity that compounds over time, not a transfer payment that vanishes once spent.

Cable infrastructure returns are typically in the order of twelve to fourteen per cent for investors, the range cited by industry research on subsea-cable investment returns.[1]

SMART-enabled cables earn additional revenue from scientific data contracts on top of telecommunications leasing, while government co-funding from climate and research budgets covers thirty to fifty per cent of the incremental sensor cost. The effective return on SMART sovereign cables is higher still.

If legislation guarantees that government data must transit sovereign infrastructure, then demand is guaranteed. Guaranteed demand means guaranteed utilisation. Guaranteed utilisation means guaranteed revenue. The investment case is not speculative. It is an asset with legislated customers.

A sovereign wealth fund, the European Investment Bank, or a coalition of national investment vehicles can invest at commercial rates because the legislation removes the demand risk.

The EU or a subset of willing member states can spend money on sovereign cables knowing that the money comes back, with returns, while simultaneously improving European defence posture, climate monitoring capability, and digital independence.

The programme has three components: new cable routes covering the transatlantic, Arctic, Mediterranean, and intra-European backbone; sovereign landing station acquisition or construction at critical termination points; and repair fleet capability. SMART sensor integration applies across all cable investments.

A significant portion of this is already committed. Denmark has allocated $468 million for its Arctic cable. Finland’s Cinia is leading the Far North Fiber at approximately €1.1 billion. Portugal’s Atlantic CAM is funded at €154 million. The Medusa cable is €342 million with EIB backing. Italy spent €700 million acquiring Sparkle for landing station sovereignty.

France spent €350 million nationalising ASN. The EU has allocated €347 million through the Connecting Europe Facility and designated €4.3 billion as priority for Arctic cables. Together, these commitments total approximately €4 billion already in motion across multiple member states. Europe is not starting from zero.

The demand guarantee is what turns each of these individual projects from a policy aspiration into a viable investment. Once legislation ensures that sovereign infrastructure has customers, every project in the pipeline becomes independently investable, whether the investor is the EU, a national government, a sovereign wealth fund, a development bank, or private infrastructure capital.

Coordination between projects is valuable but not required. The demand signal is.

Not all components require the same timeline. Landing stations can be secured immediately. New cables on established routes take three to five years. Arctic routes take seven to ten. The investment is front-loaded. The returns compound.

SMART cables as a revenue source. ASN is developing cables that integrate scientific sensors for climate monitoring, seismic detection, and tsunami warning alongside telecommunications. This technology is European. It is being developed in Trondheim, Norway, with forty years of fibre sensing heritage. No US hyperscaler cable offers equivalent capability.

The dual-use framing unlocks funding from climate, research, and disaster prevention budgets that pure telecommunications cables cannot access. But SMART is not just a funding mechanism. It is a revenue source.

Climate-resilient cable design produces real-time ocean environmental data (temperature, acidification, current patterns, seismic activity) that has commercial buyers: insurance underwriters pricing maritime risk, shipping operators routing around hazards, climate modellers calibrating predictions.

European sovereign cables should be SMART cables by default: carrying data, monitoring the ocean, and generating revenue from scientific data contracts alongside commercial leasing. If legislation mandates the procurement of SMART-enabled sovereign cables for government-funded routes, Europe makes money on improving its own defence and environmental monitoring capability.

The ownership flip. The current model is Position 4: Europe leases capacity on American-owned cables. The model should be reversed. Europe should own cables and lease capacity to hyperscalers. The hyperscalers need transatlantic and intra-European bandwidth as much as European operators do. They would lease on sovereign European cables because they need the capacity.

But the terms, including landing conditions, data protection requirements, inspection rights, and maintenance contributions, would be set by the owner, not the tenant.

A critical distinction: sovereignty is about ownership. It is acceptable, even desirable, that US hyperscalers use most of the capacity on European cables. Traffic volume does not determine sovereignty. Control does.

Why not simply negotiate better terms on existing infrastructure? Because terms are not sovereignty. Under the CLOUD Act, any data transiting US-owned infrastructure is accessible to US authorities regardless of contractual assurances.

The NSA’s UPSTREAM programme demonstrated that data is collected at precisely the interconnection points where European traffic enters American-owned facilities. Even encrypted data is at risk: the encryption layer of this series documented that traffic is being harvested now for decryption when quantum computing matures. A contract can be renegotiated. A jurisdiction cannot.

And infrastructure you do not own can be turned off.

Legislated demand. The investment case requires a demand guarantee. Whether through CADA, national legislation, or both, the requirement is the same: government data must reside on infrastructure that qualifies as Position 1 or 2 across the complete stack (cable, landing station, data centre, and cloud). France and Germany have already demonstrated this at national level.

Denmark and others could do so unilaterally tomorrow without waiting for EU-27 consensus.

One procurement rule creates customers at every layer of the stack simultaneously. European cloud providers need sovereign cables to win government contracts. Cable operators need European cloud customers to justify investment. The cascade is self-reinforcing: legislation creates demand, demand creates investment, investment creates infrastructure, infrastructure creates capability.

This is the European equivalent of how the CIA contract seeded AWS. The CIA did not subsidise AWS. It bought cloud services it actually needed. The contract created a guaranteed customer that made the infrastructure investment viable. Europe can do the same: deliberately, transparently, and at a fraction of the cost.

The United States has protected its domestic infrastructure through procurement legislation since 1933. Europe proposing an equivalent, defined by architectural sovereignty rather than nationality, is not protectionism. It is a ninety-three-year-overdue response.

The institutional model. Legislation creates the demand. Multiple delivery models can satisfy it. The Airbus consortium model (multi-government ownership with commercial operations) works for new cable routes. The EIB/Medusa model (development bank funding with commercial operators) works for Mediterranean and regional connectivity.

National acquisition (Italy’s Sparkle, France’s ASN) works for landing stations and manufacturing. Intergovernmental programmes (Denmark’s Arctic cable as defence infrastructure) work for strategic routes.

It is that the legislation makes all of them viable by removing the demand risk that currently makes sovereign cable investment unattractive compared to leasing from hyperscalers.

The Arctic route. The Far North Fiber from Europe to Japan through the Northwest Passage, and Polar Connect through the North Pole, bypass every current chokepoint: the Red Sea, the Strait of Hormuz, and the South China Sea. They pass through NATO-allied waters with almost no commercial shipping traffic.

Insurance premiums are lower because Arctic waters are not on the Lloyd’s Joint War Committee exclusion list. They are new capacity with no incumbent ownership to contest. Sovereign from birth.

A European Cable Security Fleet. Europe already operates more cable ships than any other region. ASN has seven vessels. Orange Marine has nine. But they serve commercial contracts, not a sovereign mandate. A European Cable Security Programme, modelled on the US Cable Security Fleet but properly funded, would guarantee that in a crisis European cables are repaired first.

The ships pay for themselves commercially. The sovereign retainer guarantees priority.

A coalition, not a consensus. This programme does not require agreement among twenty-seven member states. It requires a coalition of the countries that already have the assets, the ambition, and the geography. Seven countries are already acting. The missing step is not political will. It is architecture. And the coordination does not need to be imposed from above.

If legislation creates the demand, commercial incentive creates the connections. A cable operator in Portugal and a cloud provider in Finland both benefit from interconnecting if doing so qualifies both for government procurement contracts. The internet itself grew this way: through interconnection driven by commercial logic.

Legislation sets the rules. Commerce builds the network.

The Timeline

The case for urgency does not rest on abstract sovereignty principles. It rests on the permanent archive.

The cryptography problem (Paper 5): data transiting non-sovereign infrastructure is being collected and stored now for future decryption when quantum computing matures.

Every European communication, financial transaction, and intelligence message routed through American-owned cables today will become readable to American intelligence services within the next decade. Encryption that is unbreakable today is not unbreakable in 2035.

What sovereign infrastructure prevents is the archive growing larger. Every year of delay is another year of government communications, financial transactions, health records, and personal data added to a collection that cannot be retrieved. The transition timeline is not about perfection. It is about stopping the bleeding.

Phase 1, by 2027: classify the stack. Every EU institution and member state government audits and publishes the sovereignty position of each layer in its digital infrastructure: cable, landing station, data centre, cloud, application. Europe cannot fix a dependency it cannot see.

Phase 2, by 2028: legislate demand. Mandate that government critical data (defence, financial infrastructure, health, intelligence) must reside on infrastructure that qualifies as Position 1 or 2 across the complete stack. Whether through CADA, national legislation, or both. France and Germany have already demonstrated this at national level. The legislation does not build the cables.

It creates the customers. The customers create the investment case.

Phase 3, by 2033: full government procurement on Position 1 or 2. All EU institutional and member state government cloud procurement must meet full-stack sovereignty. By this point the infrastructure funded by Phase 2’s demand signal is operational. The expansion from critical data to all government data creates the scale that makes the system commercially self-sustaining.

Phase 4, by quantum: full sovereign capability. By the time quantum decryption is viable (estimates range from 2035 to 2040), Europe must have Position 1 or 2 capability for all sensitive data categories. The deadline is set by physics. The infrastructure either exists when quantum arrives or everything harvested before that point is readable.

Each phase creates the demand that funds the next. The intervention is legislation. The returns are commercial. The timeline is set by quantum.

What Follows

EllaLink went live five years ago. It proved that Europe can build sovereign cable infrastructure, land it in a sovereign facility, and operate it commercially. In the five years since, four American companies took control of ninety per cent of transatlantic cable capacity. Most of their cables land in US-owned data centres. Europe’s public backbone is ageing out.

The Baltic cables were cut. The Red Sea and Hormuz closed. And a Microsoft executive admitted under oath that the architecture cannot guarantee what it promises.

Seven countries decided independently that sovereign cable infrastructure is worth building. They are spending billions, and landing their cables in European-owned facilities. The investment case is there: twelve to fourteen per cent returns on an asset class where legislation can guarantee the customers. The money comes back.

The first step is the simplest. Ensure that every sovereign cable lands in a sovereign facility. EllaLink did this, at Sines, in a Portuguese state-owned building. It is the exception. It should be the rule. The second step is legislation: government data on sovereign infrastructure, full stack, no exceptions. The legislation creates the demand.

The demand makes every cable project in the pipeline investable. The third step is already happening. Seven countries are building. The Arctic routes bypass every chokepoint. SMART cables generate revenue from climate and defence data. The components exist. The order does not.

The infrastructure does not care about intentions. It cares about ownership. And ownership, in 2026, is not European.

[1] Subsea cable infrastructure returns historically range from twelve to fourteen per cent. See “Going deep for subsea cable investment returns”, ION Analytics Infralogic, available at https://ionanalytics.com/insights/infralogic/going-deep-for-subsea-cable-investment-returns/.