When the cloud lands
By Muad Zaki,
Director of Economic & Trade Policy at Asia Middle East Center (AMEC)
Malaysia’s data-centre boom may test not only its grid, but its politics.
The cloud has a habit of sounding lighter than it is.
Artificial intelligence is spoken of as if it floats somewhere above the economy. Cloud computing suggests something distant and weightless. Digital transformation is sold as clean, modern and almost frictionless. Yet the machinery behind this future is anything but weightless. It requires land, electricity, water, concrete, cooling systems, substations, diesel generators, security fences and roads strong enough for construction traffic.
The digital economy, in other words, has a very physical body.
Malaysia is now discovering this. Its data-centre boom has been welcomed as proof that the country is becoming a serious player in the regional technology race. Investment agencies speak of foreign direct investment, artificial intelligence, cloud infrastructure and long-term spillovers. State governments see industrial momentum. Technology companies see a useful location close to Singapore, with cheaper land and room to grow.
There is truth in all this. A country that wants to compete in the age of artificial intelligence needs the infrastructure on which artificial intelligence runs. Servers must sit somewhere. Cables must land somewhere. Power must come from somewhere.
The question is not whether data centres are useful. They are. The question is whether Malaysia is pricing their politics correctly.
Data centres have an awkward feature. Their benefits are often national, abstract and delayed. Their costs are local, visible and immediate. The gains appear in investment statistics, technology strategies and corporate announcements. The irritations appear as dust on laundry, lorries on roads, anxiety over water and questions about electricity demand.
That gap is where politics begins.
In many countries, data centres arrived first as symbols of progress. They were promoted as the infrastructure of the future: job-creating, investment-attracting and innovation-friendly. Only later did the argument change. Communities began asking not whether the facilities were modern, but whether they were worth the resources they consumed. The cloud, once a metaphor for convenience, became a neighbour.
Malaysia would be wise to notice the pattern before it becomes local habit.
The country’s enthusiasm is understandable. Data centres fit the language of the moment. They are tied to artificial intelligence, digital services, fintech, cybersecurity and cloud computing. They allow leaders to speak about the future without seeming speculative. Unlike start-ups, they involve steel, land and large capital commitments. They are easy to photograph and easier to announce.
But headline investment is a poor substitute for public value.
A factory employing thousands of people has obvious politics. A port, a logistics hub or a tourism project creates visible chains of local activity. Workers are hired. Shops benefit. Transport firms gain business. Families can see where the money goes.
A data centre is harder to sell after the ribbon is cut. During construction it may create work. Once operating, however, many facilities are relatively quiet employers for their size and cost. The capital intensity is impressive. The labour intensity is not.
This does not make them bad investments. It makes them politically delicate ones.
Voters rarely think in terms of capital expenditure. They think in terms of household consequences. A billion-ringgit project that produces few visible local jobs but consumes water and electricity is not easily defended at a community meeting. Economists may speak of spillovers. Consultants may speak of ecosystems. Voters ask whether their children were hired.
This is the first danger for Malaysian politicians: mistaking an investment story for a constituency story.
The second is mistaking technological enthusiasm for public consent.
Most Malaysians use and welcome technology. They want better internet, smoother digital services, efficient banking, online commerce and modern government. But using technology is not the same as hosting its industrial infrastructure. People may like artificial intelligence in their phones and still dislike the idea of nearby facilities drawing heavily on water, land and power.
That is not backwardness. It is rational politics.
The public rarely objects to progress in the abstract. It objects when progress appears to send the benefits elsewhere and the burdens next door.
Johor offers an early warning. Its rise as a data-centre hub has brought attention and investment. It has also brought complaints about dust, pollution and water concerns. These are not yet a national backlash. But they are the kind of small signals that often precede larger political arguments. In other countries, the debate also began locally: a road here, a substation there, a water worry, a tax incentive, a noise complaint. Then the questions accumulated.
The most combustible issue may be water.
Electricity demand can be hidden inside technical language: grids, capacity, tariffs, generation mix. Water is harder to abstract. It enters kitchens, bathrooms, farms, schools and restaurants. If households come to believe that data centres are receiving secure access while ordinary users face pressure, the facts will have to fight the feeling. In politics, feeling often arrives first.
That is why the standard defence of data centres, though partly correct, is insufficient.
Industry advocates argue that data centres should not be judged merely by direct jobs. They point instead to long-term ecosystem benefits: artificial intelligence capabilities, cloud services, cybersecurity, supplier networks, skilled work and the attraction of future industries. This may be true. But politics runs on a different clock. An ecosystem may take ten years to mature. An election may be lost in five.
This timing problem matters. The costs of a project are often experienced early: construction, land-use change, utility anxiety, public incentives. The benefits may be diffuse and delayed. A technology cluster, if it emerges, may enrich firms and skilled workers far from the affected community. That does not mean the investment lacks value. It means the value may not be felt where the political cost is paid.
America and Ireland have already shown how quickly the mood can turn. Places that once competed for data centres now debate their power demand, land use, water consumption and tax treatment. The facilities did not suddenly become less digital. Residents simply began measuring them differently. What had been sold as national competitiveness was judged as local inconvenience.
Malaysia’s leaders should not read these examples as arguments for retreat. They should read them as warnings against laziness.
The lazy approach is to count investment announcements and call that strategy. The better approach is to ask what remains after the announcement: how many permanent local jobs, how much tax after incentives, how much water, how much electricity, how much local procurement, how much training, how much infrastructure cost, and how much public trust.
A project that cannot answer these questions may still be profitable. It may even be useful. But it is not yet politically safe.
This is the uncomfortable truth of the data-centre race. The countries that win may not be those that approve the most projects fastest. They may be those that learn to discriminate. Some data centres will be worth welcoming. Others may consume too much scarce resource for too little local return. A serious industrial policy must know the difference.
Malaysia has time to get this right. Its advantage is that others have made mistakes first. The choice is not between embracing the future and rejecting it. It is between governing the future and merely advertising it.
For politicians, the lesson is sharper still. Voters do not reward megawatts, server racks or investment memoranda. They reward visible improvements in ordinary life. If data centres bring skills, jobs, infrastructure and shared prosperity, they will become part of Malaysia’s success story. If they bring anxiety over bills, water and land while the benefits remain invisible, they will become someone’s campaign problem.
The cloud may be digital. The backlash will not be.
*Disclaimer: The views and opinions expressed in this article are those of the author and do not necessarily reflect the official policy or position of any organization, institution, or group with which the author is affiliated.
By:
Director of Economic & Trade Policy
at Asia Middle East Center (AMEC)
