Is Singapore’s Data-Centre Boom at Risk? How US AI Chip Restrictions Could Change the Investment Landscape

Singapore’s data-centre ambitions face a new geopolitical challenge

For years, Singapore has been positioning itself as one of Asia’s most important data-centre and technology infrastructure hubs.

The country’s stable political environment, strong telecommunications infrastructure, international connectivity and access to financial capital have made it an attractive destination for technology companies.

The growing popularity of artificial intelligence has added another dimension to this opportunity.

AI models require enormous computing resources. These resources are provided through specialised data centres equipped with advanced processors, networking infrastructure and sophisticated cooling systems.

As demand for AI computing increases, data-centre operators could potentially enjoy higher utilisation, stronger rental demand and opportunities to expand their infrastructure.

However, a new development from Washington could complicate this investment story.

On 16 November 2026, the US Department of Commerce announced tighter export controls covering advanced semiconductors destined for Singapore, Malaysia, Indonesia, Thailand and Brazil.

The announcement has raised concerns about the ability of data-centre operators in these countries to procure advanced AI chips and provide computing services to certain overseas customers.

For Singapore investors, this raises an important question:

Could geopolitical restrictions undermine the investment case for Singapore’s data-centre industry just as AI demand is accelerating?

I believe the answer requires a closer examination of how the industry operates.

Why are data centres suddenly caught in the US-China technology rivalry?

To understand the implications, investors first need to understand how AI infrastructure works.

Traditional data centres provide computing, storage and networking services.

AI-focused data centres require considerably more specialised infrastructure.

They need powerful graphics processing units, or GPUs, capable of handling complex AI calculations.

Companies such as Nvidia have become central suppliers of these processors.

However, Washington is concerned that Chinese companies could gain access to restricted American computing technology through overseas data centres.

Instead of physically importing advanced AI chips into China, a Chinese company could potentially rent computing capacity from a data centre located in Singapore or another country.

This creates a regulatory challenge.

Restricting physical chip shipments alone may not prevent Chinese companies from accessing the computing capabilities of those chips.

The new US measures therefore introduce additional requirements concerning advanced semiconductor exports and certain overseas computing arrangements involving Chinese customers.

For data-centre operators, this means customer verification and regulatory compliance could become increasingly important.

Reuters reported that the restrictions could affect the wider Southeast Asian data-centre investment pipeline, particularly businesses serving Chinese AI customers.

The immediate concern is that licensing uncertainty could delay projects and complicate financing decisions.

Singapore’s data-centre industry: Three possible consequences

1. Data-centre expansion could become more complicated

The first potential consequence concerns the development of new facilities.

Building a modern data centre is a capital-intensive undertaking.

Operators must secure land, electricity, cooling infrastructure, networking equipment and financing.

For AI-focused facilities, access to advanced computing hardware is another important requirement.

If licensing requirements delay chip procurement, certain projects could experience longer development timelines.

This does not mean every Singapore data-centre project will be delayed.

Projects serving established international technology companies and legitimate domestic customers may have different regulatory exposure.

Nevertheless, investors should recognise that hardware availability and regulatory approval are becoming additional considerations when evaluating future expansion.

2. Customer concentration could become a more important investment risk

Historically, investors evaluating data-centre companies have focused on occupancy rates, rental growth, lease duration and tenant quality.

These factors remain important.

However, the latest US restrictions introduce another question:

Who is ultimately using the computing infrastructure?

A data-centre operator serving global cloud providers, financial institutions and established multinational companies could have a different risk profile from a business heavily dependent on Chinese AI developers.

The distinction could become particularly important for companies offering GPU-as-a-Service arrangements.

These businesses provide customers with access to powerful computing resources without requiring them to purchase the physical hardware.

While the business model could generate attractive revenue opportunities, it may also create additional regulatory obligations.

Investors should therefore examine customer concentration and the nature of computing services provided.

3. Compliance costs could affect profitability

The third consequence is less obvious but potentially important.

Companies operating advanced AI infrastructure may need to strengthen:

  • Customer identification procedures.
  • Ownership and beneficial ownership verification.
  • End-user monitoring.
  • Transaction documentation.
  • Export-control compliance.
  • Internal risk management systems.

These requirements could increase operating costs.

Larger international operators may have more resources to manage these obligations.

Smaller operators could face relatively greater compliance burdens.

Over time, regulatory compliance may become another competitive advantage for established data-centre businesses.

Which Singapore-listed companies should investors monitor?

Keppel: Can infrastructure investments deliver attractive returns?

Keppel is one of the more interesting Singapore-listed companies in this discussion.

Its exposure to infrastructure and data centres gives investors access to the broader digital infrastructure investment theme.

The company’s investment case should not be evaluated simply by looking at AI demand.

Investors also need to examine whether new infrastructure projects can generate attractive returns relative to the capital invested.

The latest US restrictions could affect certain parts of the data-centre ecosystem, depending on customer exposure and hardware requirements.

However, Keppel’s broader infrastructure portfolio means investors should avoid treating the entire group as a direct proxy for advanced AI chip exports.

Key investment considerations:

  • Data-centre investment pipeline.
  • Contracted revenue.
  • Capital expenditure commitments.
  • Asset management opportunities.
  • Returns on infrastructure investments.

Singtel: Is Nxera positioned to benefit from trusted AI demand?

Singtel’s data-centre strategy deserves particular attention.

Its Nxera platform is part of the group’s effort to expand its regional data-centre presence.

The company also has exposure to sovereign AI infrastructure.

This could become an important competitive advantage if governments and large enterprises increasingly prioritise trusted computing environments.

Following the US announcement, Singtel indicated that the expected impact on its sovereign AI business would be minimal.

It also said that its exposure to neocloud customers through Nxera was insignificant relative to contracted revenue.

This provides some reassurance, although investors should continue monitoring the company’s hardware procurement arrangements and customer profile.

The bigger question is whether Singtel can turn its AI infrastructure investments into sustainable earnings growth.

Building data centres is only the first step.

Generating sufficient returns from those assets is ultimately what matters to shareholders.

Data-centre REITs: Could regulatory uncertainty affect valuations?

Singapore investors also have exposure to data-centre infrastructure through REITs.

These investments provide access to data-centre properties without requiring investors to own and operate the facilities themselves.

However, investors should distinguish between property landlords and operators selling computing services.

A data-centre REIT typically earns rental income from tenants that occupy its facilities.

Its immediate exposure to export controls may therefore be more indirect.

Potential consequences could include:

  • Tenants delaying expansion plans.
  • Changes in demand for additional data-centre space.
  • Higher compliance-related costs for certain tenants.
  • Uncertainty surrounding future development projects.
  • Changes in investor expectations for rental growth.

For investors, the most important considerations remain tenant quality, lease structures, occupancy, debt costs and distribution sustainability.

The regulatory exposure of each REIT will depend on its underlying tenants and assets.

Could Malaysia become a stronger competitor?

An interesting question is whether Singapore could lose some of its competitive advantage to Malaysia.

Malaysia has attracted considerable interest from international technology companies seeking to establish data-centre infrastructure.

Johor, in particular, has benefited from its geographical proximity to Singapore.

The region offers opportunities for large-scale infrastructure development, although electricity, water, connectivity and regulatory considerations remain important.

If Singapore-based companies face additional regulatory uncertainty, some future investment decisions could favour alternative locations.

However, investors should not assume that Malaysia automatically becomes the beneficiary.

The US restrictions also cover Malaysia.

Consequently, Malaysian data-centre operators are not necessarily insulated from the new requirements.

The competitive landscape will depend on how the regulations are implemented, the ownership of the operators, the customers being served and the ability to secure advanced computing hardware.

Singapore’s response could determine the longer-term outcome

Singapore’s government has indicated that it is studying the new restrictions and engaging American authorities.

On 18 November, Singapore and the US agreed to follow up on practical arrangements to facilitate continued trusted access to advanced American technologies.

Prime Minister Lawrence Wong also indicated that Singapore could facilitate checks by US authorities on how advanced chips are being used locally.

This is important because Singapore’s long-term interest is not simply to preserve semiconductor trading volumes.

It is to maintain its reputation as a trusted international business and technology hub.

The government has made clear that it will not allow companies to misuse Singapore’s business environment to circumvent foreign export controls.

For legitimate companies, maintaining access to advanced technology is essential.

My investment assessment: The data-centre opportunity is evolving, not disappearing

I would be cautious about interpreting the latest US restrictions as the end of Singapore’s data-centre investment boom.

The long-term demand for computing infrastructure remains supported by several structural developments:

  1. Growing enterprise adoption of artificial intelligence.
  2. Increasing demand for cloud computing.
  3. Digitalisation across financial services and other industries.
  4. Rising requirements for data storage and processing.
  5. Greater interest in sovereign AI infrastructure.

However, the industry is becoming more complicated.

In the past, investors could largely focus on demand growth, rental rates, occupancy and capital expenditure.

Now, regulatory compliance, customer ownership and access to advanced chips are becoming additional investment considerations.

This could create a wider gap between strong and weak operators.

Companies with established customers, diversified revenue streams and the ability to navigate complex regulatory requirements may become more attractive.

Conversely, businesses relying on uncertain customer arrangements or aggressive expansion plans could face greater scrutiny.

For Singapore investors, the real question is no longer simply how much AI computing capacity the country can build. It is how profitably and sustainably that capacity can be deployed within an increasingly complicated geopolitical environment.

Stay Ahead of the Market

Want More Singapore Investment Insights?

If you found this analysis useful, there is much more to explore. The Cicada Edge brings you more finance and investment insights, helping you look beyond the headlines and understand what really matters to investors.

Subscribe to The Cicada Edge - YouTube for more investment-focused videos and market perspectives.

Connect with us and join the conversation:

Follow along for more Singapore stocks, REITs, markets and investment ideas.

Scroll to Top