Grants and start-up funds for software, AI, robotics, EV and deep tech firms in Japan, South Korea, Thailand, Australia, Singapore, Malaysia, Pakistan and India, closing from 28 October 2026 to January 2027 or open on a rolling basis.
Thailand sets data centre rules with 20 MW industrial-zone line
EditorialBy TrustList Editorial
Approvals move to a one-stop centre at the digital ministry, after more than 110 applicants asked for over 30,000 MW of power and the government paused nearly 50 projects in September.
- Thailand
- Bangkok, Thailand
- Data Center
- Artificial Intelligence - AI
- +2 more
About Thailand sets data centre rules with 20 MW industrial-zone line
Thailand sets data centre rules with 20 MW industrial-zone line
10 November 2026: Thailand's Cabinet has approved rules that sort data centres into seven classes by power draw and confine every site of 20 MW or more to industrial zones. Deputy Prime Minister and Finance Minister Ekniti Nitithanprapas announced the decision on Tuesday, after a rush of applications that the government says far exceeded genuine demand.
Not yet independently verified. This rests on reports by the Bangkok Post and AFP; the government's own site did not answer our requests. The noise and temperature figures are described as proposed, and no effective date has been published yet. We will update this when it can be confirmed, and remove this note.
Seven classes and one dividing line
The classes run from below 0.5 MW, through bands of 0.5 to 5, 5 to 10, 10 to 20, 20 to 50 and 50 to 200 MW, to sites of 200 MW and above. The line that matters is at 20 MW. Below it a data centre may be built in a commercial zone and follows environmental codes of practice, whose details will be set after a public consultation. At 20 MW and above, stricter rules apply and the site must be in an industrial zone.
According to AFP, the proposed standards cap noise at the site boundary at 70 dBA, limit the rise in nearby temperatures to between 1 and 2 degrees Celsius, and require cooling water to be recirculated and tested before it is discharged. Projects approved before 14 August get a year to meet the environmental guidelines.
One door for approvals
Until now approvals were split across several agencies, which the Bangkok Post says encouraged developers to reserve electricity capacity on speculation. The Ministry of Digital Economy and Society will run a one-stop service centre with a screening committee, and plans a dashboard, the TH-DC Dashboard, to track projects in real time.
Investment applications will be scored on jobs, skills, local supply chains and clean energy, not only on the money committed. Danucha Pichayanan of the national economic planning council said data centres are generally not required to carry out a full environmental impact assessment.
Thirty gigawatts of requests
The figures behind the decision explain its timing. More than 110 applicants have asked for over 30,000 MW of power, and Ekniti said a quarter or fewer of them are genuine. The Board of Investment has approved 57 projects, of which 17 are operating, 20 hold construction permits and 20 are preparing to invest. In September the government asked the developers of nearly 50 projects under construction to pause.
"The use of national resources must not adversely affect the public," Ekniti said, as quoted by AFP.
What it changes for capacity in Thailand
For a company planning to host workloads in Thailand, the rules mostly affect where and how fast new large sites can be built rather than existing contracts. A provider with a project above 20 MW on commercial land will need to move or redesign it, and screening at a single committee may slow approvals while the backlog of speculative requests is cleared. Our data centre management software category lists nine products.
The detailed codes of practice for sites under 20 MW, and the date the rules take effect, are still to be published.
Sources
Categories & features
- Thailand
- Bangkok, Thailand
- Data Center
- Artificial Intelligence - AI
- Energy
- Regulatory Compliance
TrustList Weekly
The week in software and IT, in one email
The news that matters to buyers, new rankings and our own research. Every Thursday, free, and easy to leave.
You may be interested in
The Melbourne start-up builds what it calls an AI operating system for energy and manufacturing sites. Hustle Fund, POV Ventures, Co-Ventures and Ventari joined the oversubscribed round.
The Paris CRM and billing provider, founded in 2016 and counting Bouygues Telecom as a customer, will target German and Italian energy suppliers and plans to double its team.
Alphast, founded in 2016, implements and runs IFS ERP for manufacturers from Tokyo, Thailand and the US. Exeo has signed a share transfer agreement and expects only a minor effect on this year's results.
Open calls for US and Canadian software, AI, health tech and energy start-ups from NSF, NIH, DOE, state funds and accelerators, closing between 2 November 2026 and April 2027, with amounts and dates as each funder states them.
The Linux kernel ESP page-cache flaw CVE-2026-46300, which CISA says can be chained to plant PHP webshells as root, is due by 31 October; VMware ESX CVE-2026-41709, which lets an administrator act unlogged, by 11 November.
A vendor left a chiller valve shut after maintenance, and the failover Exchange then buckled under the extra 5G load. IMDA found both preventable; Singtel had already paid about S$8 million in rebates.
The 28th General Conference on Weights and Measures brought the change forward from 2035, citing a 30% chance of a first negative leap second by then and the cost of preparing for one.
More on TrustList
Everything here links back to the same verified catalogue. Pick your next stop.
- More ThailandThe ranking for this subject
- CompaniesAgencies, consultancies and IT service providers, ranked by verified reviews.
- ProductsSoftware and SaaS with pricing, features, integrations and alternatives.
- AwardsAnnual recognition decided by verified reviews and an independent jury.
- LaunchesNew products and releases, voted up by the community every day.
- AI ModelsBenchmark scores and community ratings for every major model.
- RequestsBuyers describe what they need; vendors respond directly.
- PeopleReviewers, authors and makers with public profiles.
- ComparePut up to four listings side by side before you shortlist.