Google & Meta Take Control of Undersea Backbone

Submarine cables carry approximately 99 percent of all intercontinental data traffic. Not satellite, not microwave, not some alternative transmission technology that the industry is developing as a backup. Undersea fiber optic cables — roughly 1.7 million kilometers of them stretching across ocean floors worldwide — are the physical backbone of global internet connectivity, and the vast majority of them were built and are maintained by a small number of operators under arrangements that were designed for a telecom era that no longer exists. The ITU’s Second International Submarine Cable Resilience Summit, held in Porto in February 2026, made the vulnerability explicit: Red Sea incidents in 2024 disrupted approximately 25 percent of Europe-to-Asia traffic when cables were damaged in a conflict zone with no alternative routing sufficient to absorb the load. The disruption was not theoretical. It affected latency, throughput, and service availability for carriers and enterprises routing traffic between European and Asian data centers for weeks. At the same time, the ownership model for new cable builds is shifting. Google and Meta are now leading new submarine cable projects, displacing the traditional telecom consortium model that built most of the existing undersea infrastructure. For carriers, ISPs, and enterprise network operators that depend on intercontinental connectivity, the convergence of physical threat and ownership transition means that the assumptions underlying global network architecture need to be re-examined.

Red Sea Attacks Cut 25% of Europe-Asia Capacity

The Red Sea cable incidents of 2024 demonstrated what happens when a critical choke point is disrupted. Multiple submarine cables running through the Red Sea corridor — one of the primary routes connecting Europe to South and Southeast Asia — were damaged in a conflict zone, removing approximately 25 percent of available Europe-to-Asia capacity. Traffic was rerouted through alternative paths, but those paths were not designed to carry the load, resulting in increased latency and reduced throughput for the duration of the disruption.

The incident exposed two structural problems. First, the concentration of critical routes through geographic choke points — the Red Sea, the Strait of Malacca, the Luzon Strait — means that a single event in any of these locations can degrade global connectivity. Second, the cable repair infrastructure is limited. There are a small number of cable repair ships worldwide, and the timeline for deploying one to a conflict zone, completing repairs, and restoring service is measured in weeks or months, not days. The ITU summit in Porto focused extensively on both problems, with participating governments and operators discussing faster repair frameworks and alternative route planning. global network strategy practice helps carriers and enterprises assess intercontinental connectivity risk and plan resilient routing architectures.

Google and Meta Are Building Their Own Cables Now

The ownership model for submarine cables is undergoing a generational shift. Historically, undersea cables were built by consortia of telecom operators who shared construction costs and capacity. Each consortium member received a proportional allocation of the cable’s bandwidth, and excess capacity was sold to other carriers on a wholesale basis. That model produced the majority of the 1.7 million kilometers of submarine cable currently in service.

Google and Meta have fundamentally changed the economics. Both companies now lead new cable builds as primary investors rather than consortium participants, constructing private capacity that serves their own traffic needs first and offering residual capacity to other carriers on terms they control. Google’s Dunant, Equiano, and Umoja cables, and Meta’s Anjana and 2Africa projects, are examples of hyperscalers building infrastructure that was previously the exclusive domain of telecom operators. The implication for traditional carriers is that they are becoming capacity customers on infrastructure owned by their largest traffic-generating competitors — the same companies that are building private terrestrial fiber backbones and data center campuses on land.

Resilient Routing Means Diverse Paths Not Just Capacity

The ITU summit’s focus on resilience is well placed, but resilience in submarine cable architecture is not simply a matter of having more capacity. It is a matter of having diverse physical routes that do not share the same geographic choke points. A carrier that has two submarine cables connecting Europe to Asia, both running through the Red Sea, has capacity redundancy but not route diversity. When the Red Sea is disrupted, both cables are affected simultaneously.

Route diversity requires investing in alternative paths — trans-Pacific routes connecting Asia to the Americas, terrestrial routes across Central Asia, or Arctic cable projects that avoid traditional choke points entirely. Each alternative has different cost, latency, and construction timeline characteristics. The carriers that invested in route diversity before the Red Sea disruption were the ones that maintained service levels during the incident. The carriers that relied on capacity redundancy alone were the ones scrambling. For network operators making connectivity investment decisions in 2026, the calculus is straightforward: diverse physical paths are more expensive to build but far less expensive than the revenue impact of a sustained outage on a critical route.