There are more than 1.5 million kilometres of submarine communications cable in service as of early 2026, according to TeleGeography. Add the oil and gas pipelines, then the growing web of power cables linking electricity markets and carrying output from offshore wind farms to land, and you have the physical layer of the global economy sitting on the seabed with essentially nothing watching it.
Capital has noticed. CNBC reported on a scramble among defence contractors to sell governments a way to monitor infrastructure that received almost no attention for the three decades after the Cold War.
Nord Stream started the spending
The trigger is specific. In September 2022, months into Russia's full-scale invasion of Ukraine, explosions ruptured the Nord Stream gas pipelines under the Baltic. Investigators concluded the sabotage was deliberate. Damage to subsea infrastructure in the Arctic and near Taiwan followed.
"It's the changing threat environment that's brought this to the fore," said Katja Bego, senior research fellow at Chatham House, who describes the domain as neglected for long enough that there is a lot of catching up to do.
Two things then happened at the same time. Governments started treating seabed infrastructure as a security problem, and AI and autonomy got good enough to make continuous underwater monitoring look plausible. Bego calls the maritime environment the next logical domain after the aerial drone war in Ukraine showed what remotely operated systems do to a battlefield.
The money is moving accordingly. Italian shipbuilder Fincantieri announced in July it would take majority stakes in four companies for around 600 million euros, or $689 million, buying its way into underwater and surface drones, marine surveying and subsea communications. Chief executive Pierroberto Folgiero frames it as a move from the conventional underwater world into an unconventional one of smaller submarines, drones and underwater telecoms, with future surface ships acting as motherships coordinating vehicles throughout the water column. France's Thales has assembled a comparable portfolio across sonar, anti-submarine warfare and mine detection.
Germany's Euroatlas is selling an autonomous vehicle called GreyShark, and says it has signed more than 100 million euros of contracts with European navies it declines to name. The vehicles are unarmed. A hydrogen fuel-cell version is in development, designed for up to 16 weeks underwater or 8,000 nautical miles, an endurance the company has not yet demonstrated below the surface.
The sellers are the ones lowering expectations
What makes this story unusual is how little the vendors oversell it.
Verineia Codrean, Euroatlas chief strategy and partnerships officer, describes the problem as arithmetic rather than technology. Frigates, submarines and patrol aircraft are too expensive and too few to hold continuous presence over the areas that matter. "It is virtually impossible to cover the entire ocean," she said. "You just need to know which are the areas that are more interesting to have that constant visibility of what is happening there." Her picture of the future is fleets patrolling selected zones and converging when one of them notices something.
She also volunteers the obvious consequence. "What we are building, so are our non-allies or our enemies building as well." That has already produced a category called anti-AUV work, aimed at detecting and identifying other countries' autonomous vehicles, and the requirement to tell a NATO-friendly machine from an unfriendly one. Codrean stops short of predicting drone-only conflict and expects navies to mix crewed and autonomous assets, which makes interoperability the thing worth paying for.
Bego is blunter about the ceiling. "I don't think anyone, listening to NATO people, would pretend that [we] can protect a network of a million kilometres plus." On the technology itself: "This is not technology that's anywhere near the level of the drones we see in Ukraine. Progress is being made, but it's just very difficult. Tech hurdles are really quite significant." Water defeats the things that made aerial drones cheap. Visibility is poor, and radio does not propagate.
Her conclusion is the one investors should weigh. Autonomous underwater vehicles are "mostly about deterrence and signaling," useful to develop, and "not a panacea that's going to fix it."
The unresolved question is who pays
Most subsea infrastructure is owned or operated by private companies whose experience runs to routine maintenance and accidental anchor strikes, not deliberate interference. Governments are now pushing for more monitoring, sensors and deeper cable burial, all of which cost money nobody has agreed to spend.
Bego expects the answer to come from closer cooperation between governments and infrastructure owners rather than regulation. That is a polite way of saying the liability question is open, which is exactly where the commercial opportunity sits for anyone selling surveillance hardware. Russia and China are investing heavily in the same domain, and the drone investment boom has made capital unusually available. European defence budgets have already started redirecting money toward the seabed and the cables on it, and the autonomy being fitted to these hulls is the same class of onboard decision-making now turning up on cheap development boards in aerial systems.