The cleanest current public estimate is that the world now has more than 1.48 million kilometres of submarine internet cable in service. TeleGeography’s 2025 map also tracks 597 cable systems and 1,712 landings that are active or under construction, which gives a sense of the network’s scale and density.
Table of Contents
Ask how the internet works and many people still imagine satellites. The reality is more physical, more fragile, and more important. The global digital economy depends on fibre-optic lines laid across seabeds, connected at landing stations onshore, and maintained by a surprisingly small and specialised repair ecosystem. That infrastructure is not a niche technical detail. It is the backbone of cloud services, financial messaging, streaming, government communications, and everyday cross-border connectivity.
How much of the world’s traffic these cables actually carry
On the traffic question, the headline answer is stark. ITU says submarine cables carry approximately 99% of the world’s internet traffic and over 99% of international data exchange. OECD uses similarly forceful language, saying submarine cables carry over 99% of all IP data traffic worldwide. Older official shorthand often rounds that down to “over 95%” of international data and voice transfers, but the core point does not change: undersea cables carry essentially all serious long-distance digital traffic.
What does not exist, at least not as a single simple public number, is one universal live figure for total global throughput in terabits per second. TeleGeography notes that cable capacity is measured in different ways, especially potential capacity versus lit capacity, and the most detailed system-by-system capacity data are maintained as specialised research rather than a single public master tally. That is why serious sources usually answer this question in shares of traffic, not one neat real-time throughput number for the whole planet.
The scale is still extraordinary. ITU says nearly 200,000 km of new submarine cable were installed in 2024 alone, which shows that this is not static legacy infrastructure. It is expanding because data demand keeps rising and because operators, hyperscalers and governments all want more route diversity, more capacity and more control over strategic digital paths.
Why a single cable cut does not always become a disaster
This is where the public picture often goes wrong. A cable break is serious, but a single cable break does not automatically mean a nation goes offline. Subsea networks are usually designed with redundancy, and ENISA notes that many incidents do not even reach formal reporting thresholds because one cable can fail without causing a major outage. TeleGeography makes the same point more bluntly: operators spread capacity across multiple systems so traffic can be rerouted while repairs are under way.
That resilience is real, but it has limits. The real danger begins when several cables fail in the same corridor, near the same landing area, or along the same strategic route. ENISA warns that concentration on a single route and lack of diversity amplify the risk of coordinated attack or correlated accidental damage, and history shows what that can look like. In one of the most cited cases, multiple cable breaks near Alexandria in 2008 produced major slowdowns and sharp capacity losses, including reported disruptions of 70% in Egypt and 60% in India.
The real security problem is rarely one broken line. It is clustered failure. That can come from sabotage, a dragged anchor, a natural event, or a simple routing concentration that looked efficient on paper and dangerous in crisis.
What the real risks look like when damage happens
The first risk is the least dramatic and still the most common: ordinary marine activity. OECD says almost 40% of all cable disruptions result from accidental damage caused by fishing vessels and ships dragging anchors. ENISA likewise says the most common incidents are accidental and unintentional, usually caused by fishing and shipping. Natural hazards such as earthquakes, tsunamis, volcanic activity and underwater currents account for a much smaller but still meaningful share.
That matters because it changes how the threat should be understood. Historically, most breaks have not been acts of war. OECD says deliberate sabotage has been uncommon, and ENISA says the data on malicious acts remain limited. Yet governments now treat sabotage as a rising planning assumption because the consequences of even a few well-chosen attacks could be outsized. The policy shift is visible in Europe, where the Commission’s 2026 cable security package explicitly responds to increasing risks, intentional damage and sabotage.
The second risk is concentration at cable landing stations. These are where undersea cables surface and connect into terrestrial infrastructure. ENISA describes them as weak points that can be targeted through espionage, deliberate power cuts, physical sabotage, and attacks on network management systems. DGAP makes the strategic point even sharper: large landing hubs where many cables converge can become tempting targets and significant weak spots, especially when dozens of systems enter the same zone.
The third risk is cyber-physical compromise. People tend to imagine someone literally cutting a cable on the seabed. That can happen, but operators also worry about attacks on the infrastructure that powers, supervises and manages those systems on land. ENISA notes that landing stations and associated management systems are vulnerable not only to physical intrusion but also to cybersecurity threats. In other words, the dangerous point is often not the deepest ocean trench but the comparatively accessible edge where submarine infrastructure meets land-based networks.
The fourth risk is slow recovery. Repair is specialised, weather-sensitive and limited by vessel availability. Evidence given to the UK Parliament said a nearby fault in places such as the Irish Sea might involve up to two days of travel and five days of repair, while a mid-Atlantic fault could take seven days of transit plus around ten days of repair, with bad weather stretching the timeline to a month. A cable ship can only focus on one break at a time. That turns simultaneous incidents into a capacity problem, not just a technical one.
The fifth risk is attribution. A severed cable can be caused by an accident, poor seamanship, negligence, storm activity, equipment failure, or hostile action designed to look accidental. That ambiguity is strategically useful to an attacker and politically destabilising for states trying to respond. The EU’s current policy language reflects exactly that concern: it now treats rapid detection, response, recovery and formal attribution as core pillars of cable security.
Why recent policy moves matter
The cable debate has moved well beyond telecom engineering. It now sits squarely inside national security, industrial policy and geopolitical risk. In February 2026, the European Commission introduced a new Cable Security Toolbox and allocated €347 million for strategic submarine cable projects, including a specific call to strengthen repair capacity. The logic is unmistakable: resilience is not just about laying more fibre, but about shortening repair time, improving monitoring, increasing redundancy and making sabotage harder to execute and easier to prove.
The same shift is visible in operational preparedness. Reuters reported in March 2026 that the UK and Ireland agreed joint live exercises focused on undersea cable incidents after repeated regional concerns over hostile state activity and Baltic infrastructure disruptions. That is an important marker. Governments are no longer treating cable risk as an abstract infrastructure issue. They are treating it as a crisis-management problem.
What the honest answer looks like
So how many kilometres of undersea internet cable exist? More than 1.48 million km in service is the strongest current public figure. How much of total transmission do they carry? Effectively almost all international digital traffic, with authoritative estimates clustered around 99% and older shorthand often expressed as more than 95%.
And what are the security risks when cables are damaged? The answer is broader than “internet outage.” The real risks are service degradation, congestion, rerouting stress, financial and government disruption, exposure of weak landing points, long repair timelines, ambiguous attribution, and the possibility that several failures at once could overwhelm built-in redundancy. Single breaks are often manageable. Concentrated damage is where a technical incident becomes a strategic shock.
The larger lesson is easy to miss because the internet feels weightless. It is not. It is made of glass, metal, power systems, coastal buildings, repair ships, legal permissions and vulnerable routes through crowded seas. The digital world looks virtual only from the user’s side. Underneath, it is physical infrastructure under pressure.
Author:
Jan Bielik
CEO & Founder of Webiano Digital & Marketing Agency
Sources
TeleGeography Submarine Cable FAQs
Industry reference page with current figures on cable length, fault frequency, redundancy, and how cable capacity is measured.
https://www2.telegeography.com/submarine-cable-faqs-frequently-asked-questions
ITU Submarine cable resilience backgrounder
Official ITU backgrounder with current public summaries of traffic share, network growth, and resilience risks.
https://www.itu.int/en/mediacentre/backgrounders/Pages/submarine-cable-resilience.aspx
ITU Submarine Cable Resilience
Official ITU page with international data-exchange figures and resilience context for submarine cable infrastructure.
https://www.itu.int/digital-resilience/submarine-cables/
OECD Enhancing the resilience of communication networks
OECD report covering traffic-share estimates, disruption causes, and the policy case for stronger communication network resilience.
https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/05/enhancing-the-resilience-of-communication-networks_a47d78a1/d6920477-en.pdf
ENISA Subsea cables what is at stake
European Union cybersecurity agency report on landing-station vulnerabilities, redundancy limits, accidental damage patterns, and historical outage examples.
https://www.enisa.europa.eu/sites/default/files/publications/Undersea%20cables%20-%20What%20is%20a%20stake%20report.pdf
Joint Communication to strengthen the security and resilience of submarine cables
European Commission policy page outlining prevention, detection, response, recovery, and deterrence measures for cable security.
https://digital-strategy.ec.europa.eu/en/factpages/joint-communication-strengthen-security-and-resilience-submarine-cables
Commission increases submarine cable security with EU347 million investment and new toolbox
European Commission announcement of the Cable Security Toolbox and funding package aimed at resilience and repair capacity.
https://digital-strategy.ec.europa.eu/en/news/commission-increases-submarine-cable-security-eu347-million-investment-and-new-toolbox
Subsea telecommunications cables resilience and crisis preparedness
UK Parliament report with evidence on repair timelines, vessel constraints, and contingency challenges.
https://publications.parliament.uk/pa/jt5901/jtselect/jtnatsec/723/report.html
UK and Ireland to test readiness for undersea cable incidents
Reuters reporting on recent joint preparedness exercises and the growing geopolitical focus on cable security.
https://www.reuters.com/world/uk/uk-ireland-test-readiness-undersea-cable-incidents-2026-03-13/
