The Malaysian data centre landscape is in the middle of a transformation, in which the country tries to evolve from a secondary destination for its neighbour Singapore’s overflow demand, into a regional leader in high-value Artificial Intelligence (AI) infrastructure in its own right. This expansion is happening rapidly, reshaping the nation’s industrial policy and physical landscape, particularly in the states of Johor and Selangor as the country continues to prioritise capital-intensive, high-tech projects over traditional digital storage, New Straits Times writes.

The AI-first mandate

In recent years, Malaysia has formalised a move to focus on facilities that enhance national technological capabilities. In February 2026, Prime Minister Datuk Seri Anwar Ibrahim confirmed that the federal government had effectively halted approvals for non-AI-related data centre proposals in 2024.

This strategy is intended to ensure that limited resources such as land, power, and water are reserved for projects that offer the deepest technological benefits. However, this plan also raises the stakes and challenges the industry. As a result of this mandate, Malaysia’s focus on development has become almost entirely AI-focused. An industry estimation now sees the country as pointing towards 4.6GW of capacity currently planned or under construction. This choice of shift is supported by major international investments, in the form of commitments exceeding $2bn each that came from global giants Microsoft and Google, all issued under the MyDigital initiative.

There are also firms such as YTL Power collaborating with global AI behemoth Nvidia to deploy advanced Blackwell architecture. This has helped cement Malaysia's position at the forefront of the AI infrastructure race but it is an environment with barriers to entry, favouring well-funded, established global operators and private equity-backed platforms over smaller entrants.

The establishment of an AI-filtered sector has thus resulted in a national economic boom, and Malaysia is quickly catching up in addressing its sustainability issues. To this end, the Malaysian government has adopted a sustainability-first approach to manage the immense resource demands of these facilities.

Data centres are notoriously resource-intensive. For instance, Tier 1 and Tier 2 projects can require up to 50mn litres of water daily, which is roughly 200 times the volume of Tier 3 or 4 sites. As a consequence, the government is now rejecting nearly 30% of data centre proposals that fail to meet strict environmental standards regarding water and electricity usage.

The influx of industrial demand in Malaysia has placed a significant strain on local infrastructure and communities, as reported by Channel News Asia. In Johor’s Gelang Patah, a 15-hectare construction site for firms such as ZData and NTT Data Group is located just 150 metres from residential areas, leading to complaints about heavy dust pollution and breathing difficulties among locals.

For residents, concerns over rising electricity costs and potential water disruptions mounted and led to calls for a more structured digital zoning framework to separate industrial hubs from residential areas. Some firms respond by using reclaimed water and installing dust-prevention netting. However, experts argue that better state-level coordination of electricity and water treatment is essential before the system reaches a breaking point.

Regional impact

Malaysia’s decision to pivot toward a high-density AI workload has also caused a division in the Southeast Asian market. Kuala Lumpur has managed to capture the high-value AI frontier so far, but its regional peers are absorbing the displaced demand for traditional enterprise workloads and non-AI applications.

As a result, neighbouring markets such as Indonesia, Thailand, and the Philippines are now the primary destinations for non-AI workloads and these countries are still benefiting from the saturation and infrastructure bottlenecks in Malaysia and Singapore. In Thailand, the country has introduced renewable power reforms, while Indonesia is leveraging tax holidays and special economic zones, such as Batam’s Nongsa Digital Park, to attract hyperscalers.

Fitch Solutions’ BMI notes that Malaysia's demand has matured as its narrative is no longer merely an overspill of Singapore’s market. Malaysia has now become a primary driver of regional digital transformation. It has moved from a quantity over quality approach to one that discriminates between facilities that simply arbitrage utility costs and those that deepen the country’s position in the global digital value chain.