Undersea Cables Are the Internet’s Most Vulnerable Infrastructure — and Nobody’s Guarding Them

There are approximately 550 active undersea cable systems carrying roughly 95% of all international data traffic. They are, collectively, the physical backbone of the global internet. They are also almost entirely undefended.

The cables are not secret. Their routes are published, charted, and in many cases visible on publicly available maps. They land at coastal facilities called cable stations, which are identifiable by their distinctive antenna arrays and the heavy power infrastructure required to amplify signals across transoceanic distances. The cables themselves run along the seafloor at depths that vary from a few metres in coastal approaches to several kilometres in the open ocean.

Cutting a single major cable causes disruption but not catastrophe — traffic reroutes through surviving cables, typically with reduced capacity and increased latency. Cutting multiple cables simultaneously or in rapid succession is a different matter. A coordinated attack on the six cables that carry the majority of transatlantic internet traffic would cause significant internet outages across North America and Europe with consequences for financial markets, communications infrastructure, and emergency services that are difficult to fully model.

The Recent History

Between 2022 and 2025, four undersea cables were damaged in circumstances that remain officially unexplained. The Baltic Sea incidents of 2023 and 2024, which disrupted cables connecting Finland, Germany, and the Baltic states, were attributed by some governments to Russian interference and officially logged as “accidents” involving ship anchors. The technical evidence for anchor damage was present in some cases. Independent analysts found it unconvincing in others.

The Defence Problem

Defending undersea cables is genuinely difficult. The ocean is large. The cables are long. Naval patrol of the full route of a major cable is not feasible. Detection of cable cutting is possible but slow — it takes hours to confirm which cable has been damaged and where, and longer to identify the responsible party.

The most viable defences are redundancy, rapid repair capacity, and deterrence. The first is improving but expensive. The second requires cable repair ships that are in short supply globally and contracted through commercial arrangements that may not respond quickly to military priorities. The third requires attribution capability that currently doesn’t exist for subsea incidents.