Before the Servers Arrive: Why Sabah’s AI Data Centre Should Trigger a Manpower Policy

By Brendon Beliku, Chartered Member of the Chartered Institute of Logistics and Transport (CILT) Malaysia, writing independently on Sabah’s economic governance, labour policy, logistics, transport and regulatory affairs

KOTA KINABALU: This week, Memoranda of Understanding were signed for Sabah’s first Artificial Intelligence Data Centre, proposed by North AIDC Sdn Bhd at KKIP. 

The project’s published concept describes a 130 megawatt, liquid cooled hyperscale facility, built in three phases of 30, 50 and 50 megawatts and powered by natural gas. 

Chief Minister Datuk Seri Hajiji Noor, in a speech delivered on his behalf, was careful to note that an MoU “is not, by itself, a government approval,” and that any proposal must pass scrutiny on power, water, cooling, land and connectivity.

He also made a point that deserves more attention than the megawatts. Once operational, he said, modern data centres “generally employ comparatively small, highly skilled teams.

” That sentence is a policy cue. An investment that brings large capital but few, highly specialised jobs tells Sabah exactly where its manpower policy must go next.

-Few Jobs, High Skills, Clear Signal-

Unlike plantations or construction, data centres do not create jobs in large numbers. 

Their value lies in the skills they require: electrical and cooling engineers, network specialists, data and cybersecurity professionals, and facility managers. 

If those skills are not available locally when the facility opens, the roles will be filled from outside, and the local share of the benefit will be limited to construction and support services.

This is a reason to prepare, not to turn such investment away. The Chief Minister himself called for local procurement, specialist training, technology transfer, apprenticeships and partnerships with universities and technical institutions. 

The question is whether those intentions remain encouragement, or become policy for every investment of this type.

Sabah is better placed than most states to answer that question. Its separate Sabah Work Pass process, where local hiring priority must be demonstrated first where it applies, gives the state a lever over specialist hiring that few other states hold. 

The AI data centre is an opportunity to use that lever with intent rather than by default.

-Four Policies the Cue Points To-

First, a state manpower policy for high value investment. Every major project should submit a workforce plan alongside its technical and financial proposals, setting out the roles needed, the expected local share and how the gap will close. 

Power, water and land are already assessed before approval. Skills should be assessed in the same way.

Second, targeted automation in transport and logistics. Automation is often feared as a threat to jobs. In Sabah’s logistics sector, it is better seen as a response to shortages and delays. 

Sepanggar Bay Container Port is expanding, a vehicle booking system is being introduced, and 24-hour operations have been approved, pending readiness. 

These systems need fewer manual handlers but more trained operators and technicians. 

A targeted automation policy for ports, warehousing and freight would raise productivity while creating the kind of skilled local roles that an AI data centre itself will demand. The two agendas reinforce each other.

Third, stringent and time bound quotas for imported specialists. Specialist hiring from abroad is legitimate when skills are scarce, and an AI facility will need it at the start. 

But each approval should come with conditions: a defined quota for foreign specialists, linked to the project’s phases, and a requirement that the quota falls as local capability grows. 

A 130 megawatt facility built in three phases provides a natural timetable. The foreign share in phase one need not be the foreign share in phase three.

Fourth, succession planning that is enforced, not filed. Federal rules already require succession plans for many specialist roles under the revised Employment Pass policy. 

In practice, a succession plan is only as strong as the local successor behind it. For high value investments, each foreign specialist role should be paired with a named local trainee and a target date for transfer, with progress reported to approving authorities.

-Making the Cue Count-

Calibration matters. Conditions that are too rigid could deter the investment Sabah is seeking. 

Conditions that are too loose would leave the state with modern infrastructure and few local careers attached to it. 

The balance lies in clarity. Investors generally accept requirements that are predictable, published and applied equally; uncertainty is what discourages them.

The Chief Minister noted that Sabah can learn from Johor, where data centre growth has raised questions about utilities and nearby communities. Sabah can learn on the workforce side too, before the first server arrives.

He also offered the most useful test for any proposal: its value should be judged “by the benefits it can demonstrate and sustain, not by the scale of an announcement.

” For an AI data centre, the most lasting of those benefits will not be measured in megawatts. It will be measured in how many of the people running it are Sabahans, and how soon.

[Disclaimer]: The views, opinions and analyses expressed in this article are solely those of the author and do not necessarily reflect the views, policies or positions of the Jesselton Times, its editors, management or affiliated entities. This article is published for analytical and public discussion purposes and should not be construed as institutional commentary or endorsement. The views expressed are the author’s own and not those of the Chartered Institute of Logistics and Transport (CILT) Malaysia.

Related Articles

253FansLike

Latest Articles