This article first appeared in Forum, The Edge Malaysia Weekly on July 13, 2026 - July 19, 2026

In 1901, Kaiser Wilhelm II declared that despite a deficient fleet, Germany had “conquered for ourselves a place in the sun”. This rhetoric fuelled a hyper-competitive naval arms race aimed at securing colonial dominance; a pursuit that ultimately dragged Europe into World War I and its manifold consequences. Today, Asean finds itself scrambling for a digital version of that same “place in the sun”.

Malaysia has fast become one of Southeast Asia’s premier data centre hubs, leveraged by strategic geography, robust fibre connectivity, ample and cheap industrial land, stable power access and a deeply entrenched semiconductor ecosystem. The economic promises are familiar: modernisation, job creation, global competitiveness and unprecedented productivity. In a modern economy, digital infrastructure is as foundational as deepwater ports were to the mercantile era.

But before the next celebratory ribbon-cutting ceremony, we must pause. Unchecked infrastructure expansion carries significant economic, environmental and extractive risks in this, the era of the Anthropocene, where human behaviour is shaping nature’s response. To ensure this boom yields sustainable human flourishing rather than building an “artificial intelligence (AI) bubble”, policymakers must address three critical questions.

First, what is our blueprint: Are we misinterpreting the playbook of leading tech nations? History shows successful countries did not achieve greatness because of AI or data centres; they succeeded because they built world-class universities, encouraged local entrepreneurship, invested heavily in foundational research and development, and harnessed deep, meritocratic talent pools. China, Singapore and the Nordic countries became economic powerhouses through decades of investment in manufacturing, telecommunications, digital infrastructure and research institutions. AI is the child of these systems, not the parent. Compute capacity is an empty commodity without this foundational investment in human capital.

Second, what is the net return? Beyond headline investment numbers, is our trajectory structured to be pro-worker, pro-environment and pro-social? Without clear guardrails, we risk importing resource-heavy, low-employment infrastructure while neglecting to ask whether these tools are being designed to actively augment local wages, protect our grid and water security and remain safe for our children and students.

Job insecurity due to AI is at the forefront of many workers’ minds in Asean. The prevailing corporate narrative promises that AI will effortlessly automate away the drudgery, freeing us for high-value creative tasks. But in developing economies, this shift often means structural displacement without a safety net. Hyperscale data centres are notoriously low-employment facilities once construction ends, meaning we are dedicating massive national resources to an industry that yields very few long-term, high-wage local jobs. If our place in the sun merely means supplying cheap labour for data-annotation or outsourced content moderation under the guise of tech jobs, we have failed.

To make AI genuinely pro-worker, industrial policy must mandate that technology augments rather than substitutes human capability. This means tying data centre approvals to measurable domestic spillovers: co-investment in local talent upskilling; tax incentives linked to wage appreciation rather than just headcount; and pathways that integrate Malaysian engineers into the core algorithmic design, not just the physical upkeep.

We, the end-users, perceive AI as frictionless. We type a prompt and an answer immediately appears. A single query is trivial, consuming about as much energy as a few seconds of video streaming. The real issue is scale. AI systems handle billions of prompts daily, each served by a vast infrastructure of servers, chips and power systems running around the clock. Multiplied across that volume, negligible individual costs aggregate into enormous demand for electricity and water, straining grids and cooling systems, particularly in hot climates.

Globally, data centre electricity consumption is projected to more than double from 448 terrawatt-hours in 2025 to 945tw-hours by 2030, surpassing Japan’s total current power usage, while generating 399 million tonnes of carbon dioxide, consuming 9.3 trillion litres of water, and occupying more than 14,500 sq km. This intersecting energy, water and industrial policy crisis is highly visible in Malaysia, where 

RM144.4 billion in digital investments between 2021 and mid-2025 will push data centre energy demand past 5,000mw by 2035 — nearly 40% of Peninsular Malaysia’s current capacity — already forcing restrictions on non-AI facilities to ease grid pressure. Ultimately, the central challenge for Malaysia and the Asean region is a meaningful policy shift: determining whether new compute infrastructure will be approved only when its collective impacts on power, cooling, land, water and emissions are fully measured, priced and aligned with Paris Agreement climate commitments.

Third, how are we ensuring our geopolitical alignment? How do we align with our Asean and Global Majority neighbours to prevent a race to the bottom, ensuring we dictate our digital future rather than merely host advanced digital extraction? History warns us of the cost: In the late 19th century, Malayan getah percha (latex) was aggressively extracted by arduous manual labour to insulate the British Empire’s telegraph network, felling trees for a physical system that built Western wealth while leaving our ecosystems scarred. Today, modern fibre-optic lines trace those same colonial routes, with roughly 98% of global data traffic controlled by dominant Western tech conglomerates connecting private facilities.

Peers across the Global Majority are already rethinking policy to protect their data sovereignty. Singapore ended its data centre expansion through a strict moratorium, mandating tropical standards that force operators to run efficiently at higher temperatures to protect the national grid. Morocco has repositioned itself as a gateway to Africa by securing major AI investments in the Dakhla-Oued Eddahab region under supply contracts powered entirely by its sovereign wind and solar power. Latin American nations have banded together to finance Latam-GPT, resisting Western epistemic bias alongside regional coalitions that advocate for Indigenous data sovereignty and legal personhood for overstressed watersheds.

By rushing to host these networks without similar regional coherence, we will find ourselves in a new era of extractivism: subsidising Westernised models and their inbuilt biases that erase local oral traditions, inheriting low-wage data-labelling, and hiding environmental impacts behind opaque non-disclosure agreements. True alignment means treating our shared natural and digital assets as sovereign capital, not cheap commodities.

Yes, with certain caveats. The proliferation of data centres and AI is justifiable only if they elevate Malaysia’s national capability, resilience and skills while actively regenerating the ecological and social systems that sustain economic growth. Compute must not be treated as something to mine and ship out. It should deliver better public services, stronger local businesses and cities built for a hotter future, with benefits that people can actually see and share.

Malaysia must push back against the “Big High Tech” story of unfettered need for land, power and water, and be a true guardian of the health of people and planet. Without firm rules built on those values, the country risks watching its economy and environment spiral out of control; as W B Yeats put it, “the centre cannot hold”.

To keep this rapidly developing series of events under control, Malaysia can adopt “frugal AI”, a lean approach inspired by the French Standardisation Association. The rule is simple: AI is used only when it does more social and ecological good than a less power-hungry alternative. Every application would be audited and ranked by its footprint. Wasteful uses would be swapped for low-tech tools; vital but heavy uses, such as health or education, would be shared fairly; only the lightest technologies would be free for all.

This “Low Little Tech” approach turns AI from a high-emissions luxury into a tightly regulated public utility. Just as a green transition runs on bicycles and good public transport rather than sports cars, putting AI on the right side of history means keeping it within planetary boundaries and respectful of human rights.

Professor Tan Sri Dr Jemilah Mahmood, a physician and experienced crisis leader, is executive director of the Sunway Centre for Planetary Health at Sunway University, Malaysia. She is founder of MERCY Malaysia and has served in leadership roles internationally with the United Nations and Red Cross for the last decade. She also sits on several corporate and non-profit boards globally and nationally. In 2019, she was the first Malaysian recipient of the Asean prize.

Save by subscribing to us for your print and/or digital copy.

P/S: The Edge is also available on Apple's App Store and Android's Google Play.