Thailand’s ambitions to expand its data centre sector have hit an unexpected pause.
The move does not signal a retreat from the digital economy. Instead, the rapid pace of expansion has prompted the government to reassess the infrastructure, resources and regulatory framework needed to support the industry.
On September 4, Thailand’s Data Centre Business Policy Committee placed construction on 49 data centre projects on hold and paused approval decisions for another 117 proposed facilities. The government said the move would allow relevant agencies to establish nationwide standards covering electricity and water use, site selection, safety, environmental protection and the economic benefits of the industry.
The government expects to produce a regulatory draft within about a month after September 11, but has not specified how long the pause itself will last. Officials have stressed that the policy is not intended to shut the door on data centre investment, but to introduce a more coordinated framework for the sector.
In the first six months of 2026, Thailand’s Board of Investment approved 88 projects related to artificial intelligence and data centres, representing approximately 886 billion baht in investment.
These figures highlight both the scale of Thailand’s digital ambitions and the physical resources required to support them.
Behind every AI model, cloud service and digital platform are physical facilities that require electricity, cooling systems, land and supporting infrastructure. As Thailand seeks to build a stronger foundation for an AI-driven economy, it faces a question that extends beyond investment:
Can its infrastructure and resources keep pace with the technology?
AI Does Not Feel Weightless
For most people, AI feels almost weightless. A question is typed, an answer appears, and the interaction is over.
Behind that seemingly effortless experience, however, are servers housed in physical data centres. These facilities consume electricity to run processors, storage and networking equipment, as well as the cooling and other systems required to keep them operating.
The growing adoption of AI is changing the scale of that demand.
The International Energy Agency (IEA) estimates that data centres consumed around 415 terawatt-hours (TWh) of electricity globally in 2024, equivalent to about 1.5% of global electricity consumption. In its base-case projection, data centre electricity consumption is expected to more than double to around 945 TWh by 2030, with AI identified as the most important driver of that increase.
The global share remains relatively modest, but the local impact can be much greater.
Data centres tend to be concentrated geographically, meaning that a rapid increase in demand in one region can place significant pressure on local electricity networks even when a country has sufficient generation capacity overall. The IEA notes that this concentration can make integrating data centres into electricity grids particularly challenging.
Digital Growth Meets Physical Limits
Electricity, however, is only part of the equation.
Data centres generate considerable heat and require cooling systems to maintain operating temperatures. Depending on the cooling technology, facility design and local climate, some systems can also require significant amounts of water.
This is one reason Thailand’s new regulatory framework is looking beyond electricity demand. The government has said the standards will consider resource use, locations, safety, environmental impacts and the broader economic benefits of data centre projects.
The challenge reflects a wider issue facing countries competing for AI and data centre investment.
Attracting these facilities can bring substantial economic benefits, including capital investment, construction activity, jobs, digital infrastructure and opportunities to attract related industries.
But infrastructure cannot always expand at the same speed as investment.
Electricity grids require planning and upgrades. Water resources have finite capacity. New transmission infrastructure and renewable energy projects can take years to develop, while data centre operators can move comparatively quickly to expand computing capacity.
This creates a potential mismatch between the speed of technological investment and the time required to build the systems that support it.
For Thailand, the question is therefore no longer simply whether it can attract more data centres.
It is whether the infrastructure supporting them can keep up.
Growth With New Rules
Thailand’s decision to pause and reassess the sector is therefore less a rejection of the data centre boom than an attempt to establish clearer rules for its next phase.
The government has said it wants data centre investment to generate greater value for the Thai economy, including technology transfer, human capital development and stronger links to related industries. Prime Minister Anutin Charnvirakul has also stressed that Thailand should not simply serve as a location for data centres but should ensure that such investments contribute to the country’s long-term competitiveness.
For Thailand, the challenge is not necessarily choosing between technological growth and resource conservation.
It is finding a way to make the two develop together.
For the average user, AI may become increasingly invisible-embedded into search engines, workplaces, entertainment and everyday services.
But Thailand’s data centre pause offers a reminder of what remains out of sight.
Behind every cloud service and AI model is a physical network of servers, power systems, cooling infrastructure and, in some cases, significant demands on water and land.
The future of AI may well be digital.
But the resources it relies on are anything but.
