The Sarawak Data Centre Question

An Opportunity or a Resource Giveaway?

When the last issue (87) of The Sarawakian carried news of the RM500 million AI data centre deal, it was deemed as a vote of confidence in Sarawak’s digital ambitions. But one letter to the editor, from a Concerned Reader, urged caution, noting that the headline’s triumphant tone may have outshone the need for a closer look at the numbers. The points raised were thoughtful and well researched, and deserve a considered reply.

The jobs creation story that isn’t

The Concerned Reader’s case rested largely on evidence from the United States, where the data centre industry is most mature. American data centres account for less than 0.01% of total employment while consuming over 4% of national electricity. In Virginia, which hosts roughly a third of US capacity, the investment required to create one permanent data centre job has been calculated at nearly a hundred times the cost of creating a job in any other industry. Construction crews are real and visible, but they typically depart within two years, leaving a large facility to be operated by a few dozen specialists, often imported from elsewhere.

Energy and water demand at scale

The resource consumption figures were equally pointed. A medium-sized data centre can consume up to 110 million gallons of water annually for cooling, the equivalent of roughly a thousand households. A hyperscale facility can draw five million gallons a day. Training a single large AI model has been estimated to require a continuous power draw of 25 megawatts. Combined US data centre energy demand is expected to nearly double between 2025 and 2028, the equivalent of adding a country the size of Spain to the grid in three years. In parts of America, electricity prices have jumped by up to 20% in a single summer as data centres outbid households for grid capacity.

How disciplined jurisdictions have responded

The letter pointed to how more thoughtful governments have managed this. Singapore imposed an outright moratorium on new data centres in 2019, when the sector already consumed 7% of its electricity, and only reopened in 2022 with stringent conditions: at least 50% green energy sourcing, the most demanding efficiency targets in Asia-Pacific, and demonstrable economic contribution including job creation and technology transfer. The European Commission is pushing for carbon neutrality by 2030. Twenty-seven US states are debating whether developers should bear the full cost of new energy infrastructure their facilities require, rather than pushing those costs onto ordinary ratepayers. Governments that opened their doors without conditions, the Concerned Reader argued, are now paying for it.

Sovereign cloud versus commercial campus

The letter was not against digital investment as such. One or two strategically designed data centres for sovereign cloud purposes, securing Sarawak’s own government data and public services, would be a legitimate public interest investment. But a commercial AI campus aimed at reducing compute costs across Southeast Asia is a different proposition. The benefit flows to the investor and to the regional technology firms that buy the compute. The cost, in energy, water and grid pressure, falls on the Sarawakian public.

Why Sarawak is a natural fit

In responding, my intention is not to defend the specifics of the deal, which remain largely outside public view, but to set out why Sarawak’s particular endowments make it a plausible host for this kind of infrastructure. Sarawak’s economy is anchored in natural resources, with energy as a defining advantage. Roughly 70% of state electricity is generated from hydroelectric and renewable sources, making it both relatively clean and cost-competitive. Prices are among the lowest in the region, and Sarawak runs a surplus, already exporting power to Sabah and West Kalimantan, with exports to Brunei, Philippines and Singapore planned. These are precisely the conditions that attract data centres.

The investment also fits the direction of PCDS 2030 and the Sarawak Digital Economy Blueprint 2030, both of which prioritise digital infrastructure and industrial upgrading. With a population of around 2.5 million and a dependency ratio of 0.3, Sarawak cannot rely solely on labour-intensive growth. Some sectors will need to generate value without large workforces, and data centres are one such category. Limited job creation is not, on its own, a reason to reject an investment. The more relevant question is whether higher-value uses of energy are being crowded out.

The discipline that must follow

None of this releases the state from its responsibility to manage these resource advantages with discipline. Cheap renewable energy and surplus capacity explain why this investment is happening, and they should be treated as strategic assets rather than commodities sold off below their true value. Electricity must be priced to reflect that value. Water use deserves equal scrutiny. At the system level, a data centre creates the kind of stable, long-term electricity demand that supports further investment in renewable capacity and reinforces Sarawak’s emerging role as a regional energy hub.

A data centre is not a digital economy

But a data centre is not enough, on its own, build a digital economy. It certainly will not attract knowledge workers, nor produce the institutional depth that genuine transformation requires. What it can do is generate predictable revenue and anchor demand. Whether that revenue is eventually channelled into education, institutions and higher-complexity industries is the test that matters.

A conversation worth continuing

I am grateful to the Concerned Reader for prompting this exchange. The arguments raised were considered and evidence-based, and the kind of scrutiny they invite is exactly what a young economy needs as it weighs decisions of this scale. Sarawak’s natural endowments are finite, and the choices being made now will shape the state for decades. The question of how many data centres Sarawak should host in the long run is a separate calculation, and one well worth returning to soon enough.

Mooreyameen Mohamad