MANILA, 20 AUG 2026 — The Philippines' three largest telecommunications operators have jointly proposed a domestic submarine cable network costing about US$500 million, running from Batanes to Palawan with branches into the Visayas and Mindanao. They say it can be built within 24 months of approval.
PLDT, Globe and Converge compete for the same subscribers. Forming a consortium for domestic backhaul means the economics of building it alone no longer work for any of them.
The proposal
The plan was presented to the Department of Information and Communications Technology on 3 August. It includes connections to international landing stations, adding capacity for traffic moving within the country and overseas.
The design includes seismic sensors to detect earthquakes and tsunamis, feeding real-time data for disaster preparedness and, reportedly, for national-defence applications.
An archipelago is the hardest connectivity problem there is
The Philippines has more than 7,000 islands, and that single fact explains most of what is frustrating about internet service there.
Terrestrial fibre works well within an island and stops at the coast. Every crossing between islands needs either a subsea cable, a microwave link with limited capacity, or satellite. Microwave and satellite are cheaper to deploy and worse in every other respect, and they are what a great deal of inter-island traffic still runs on.
The consequence is that a user in Cebu talking to a service hosted in Manila may traverse a chain of hops with capacity constraints at each one, and a typhoon that takes out a single microwave tower can isolate a province. National connectivity in an archipelago is not a coverage problem; it is a topology problem, and topology is expensive.
There is a funding structure question buried in the phrase "presented to the department". Domestic backbone of this kind is built three ways: entirely by private operators recovering the cost through wholesale pricing, by government with operators as customers, or by some blend with a public guarantee. The final structure will determine who bears the risk of inaccurate traffic forecasts and whether the cable is regulated as essential infrastructure or treated as private property. Nothing published says which structure is proposed.
Why three competitors are doing this together
Subsea cable is the part of telecommunications infrastructure with the worst returns for a single operator and the best returns for a country.
The capital cost does not scale down. Laying a cable to Palawan costs roughly the same whether one operator's traffic runs over it or three do, and no single Philippine operator has enough domestic inter-island traffic to justify the spend against its own subscriber base alone. Split three ways against three sets of traffic, the arithmetic changes.
Consortium building is standard practice on international routes for exactly this reason, and it is unusual domestically because operators normally compete on network quality inside their home market. The decision to share infrastructure signals a mature market: inter-island backhaul is now seen as a common cost, not a competitive differentiator.
The risk in that arrangement is the one every shared-infrastructure deal carries. Three competitors jointly owning the backbone every one of them depends on creates a body with considerable power over pricing and access for anyone outside it, and the terms on which a fourth operator or a regional ISP could buy capacity are exactly what regulators should be examining.
The sensors are not a footnote
Adding seismic instrumentation to a telecommunications cable is a good, underused idea.
The cable is being laid anyway, along routes that cross seismically active seafloor, with power and a data path already present. Adding sensors is marginal cost on a project that is happening regardless, and the Philippines sits on the Pacific Ring of Fire with a coastline exposed to tsunami risk that current instrumentation covers unevenly.
The caveat is that dual-use infrastructure brings dual-use governance. A network that detects seafloor activity for disaster warning is also a network that observes seafloor activity, and the proposal reportedly notes national-defence applications. Who reads that data, under what oversight, and whether the disaster-warning purpose stays primary are questions worth asking before the cable is in the water rather than after.
What it would mean if built
The direct benefit is capacity and resilience: more inter-island bandwidth, and a ring topology in which a single break degrades service rather than severing a region.
The indirect economic benefit is more significant. Reliable inter-island capacity makes it possible to serve customers from local data centres and to build businesses in secondary cities that depend on continuous connectivity. We reported on Skyro reaching break-even on consumer lending distributed across roughly 10,000 retail locations, and on the central bank's digital payments trajectory. Both depend on connectivity reaching places the fibre currently does not.
It also changes where infrastructure can sit. We wrote today about data centres moving away from hub cities in search of power. That option only exists where the network reaches the new location, and in an archipelago the network is the constraint before the grid is.
Compare the ambition honestly, too. Half a billion dollars is a serious domestic commitment and it is roughly a third of what a single Johor data centre campus raised in green financing this month. Connectivity infrastructure is cheap next to compute. A national digital strategy that leads with data centres often overlooks this: the cable to a province costs less than the server building, and without the cable, the building serves nobody outside the capital.
What we could not establish
It is not yet known whether the government will approve the project, or on what terms. This is a proposal presented to a department, not a signed project, and the regulatory conditions attached will determine whether it is built and who gets to use it.
Also unestablished: the capacity of the system in design terms; how the US$500 million splits between the three operators and what governance the consortium would have; the wholesale terms on which non-members could buy capacity; whether any public funding or guarantee is sought; the landing station locations; how the seismic data would be governed and who would access it; and whether the 24-month construction estimate accounts for permitting, which is usually the longest phase.
What to watch
Watch the access terms above everything else. A shared national backbone owned by the three incumbents is either a public good or a barrier depending entirely on what a non-owner pays for capacity, and that will be settled in regulatory conditions rather than in the announcement.
Then watch the timeline against permitting. Twenty-four months of construction is credible; twenty-four months from approval to service depends on how quickly rights of way, environmental clearances and landing permissions move, and infrastructure projects in the region routinely lose a year there.
Finally, watch whether other archipelagic states copy the structure. Indonesia has the same topology problem at greater scale, and a domestic consortium model that works in the Philippines would be a template with an obvious next customer.
