- 6 hours ago
Malaysia has become one of Southeast Asia's fastest-growing destinations for AI and data centre investment. The challenge now is ensuring billions in investment translate into substantive value for Malaysians.
From financing and tax policy to energy security, climate resilience and digital competitiveness, the next phase will be defined by the quality of decisions made today.
On this week's NIAGA SPOTLIGHT, Tehmina Kaoosji speaks with Dr Jasmine Begum, General Manager, Corporate External Legal Affairs, Microsoft Asean, together with Gan Pei Tze, International Tax and M&A Head, PwC Malaysia, about what it will take to build an AI economy that delivers lasting value for Malaysia.
From financing and tax policy to energy security, climate resilience and digital competitiveness, the next phase will be defined by the quality of decisions made today.
On this week's NIAGA SPOTLIGHT, Tehmina Kaoosji speaks with Dr Jasmine Begum, General Manager, Corporate External Legal Affairs, Microsoft Asean, together with Gan Pei Tze, International Tax and M&A Head, PwC Malaysia, about what it will take to build an AI economy that delivers lasting value for Malaysia.
Category
đź—ž
NewsTranscript
00:01Hello, I'm Tamina Khosji and this is Nyaga Spotlight.
00:04Now, today on analysis, what we are going to be going through is actually Malaysia's AI economy challenges,
00:10including investment infrastructure and climate.
00:13Malaysia has been attracting billions of ringgit in AI and data centre investments,
00:17positioning itself as one of our Southeast Asia's investment hubs.
00:21However, we also have climate pressures growing on the horizon,
00:25together with countries competing for the next wave of digital infrastructure.
00:29This focus now shifts to, of course, tangible value for everyday Malaysians.
00:34Are we building an AI economy that creates jobs, sustains innovation and digital impact,
00:39or are we simply expanding infrastructure that also risks becoming outdated rather quickly?
00:45Well, today we're going to be examining the whole breadth and scope of all of these issues
00:49and what it will take for Malaysia to stay competitive in an AI race that is changing faster than ever.
00:56Welcoming to the studios this morning, Dr. Jasmine Begum,
00:59who is General Manager with Corporate External Legal Affairs for Microsoft ASEAN,
01:04together with Gan Pesa, International Tax and M&A Tax Head with PwC Malaysia.
01:10A very good morning to you ladies. Thank you so much for being with us here.
01:12So, Dr. Jasmine, if I can get started off by going back to that big number,
01:19which is, of course, the billions in AI and data centre investment,
01:23which Malaysia has been attracting over the past two years alone.
01:26Let's not look so far back in the past.
01:29Now, every government naturally wants to become an AI hub.
01:33It is the technology of the moment.
01:35Now, from Microsoft's work across ASEAN,
01:38what would you say distinguish countries that are able to attract the investment
01:43versus those which are able to ensure there is sustained value over the long term?
01:51Thank you, Tamina.
01:52There are two things that I always talk about there.
01:56Enabling regulatory environment.
01:58We need a progressive, digital-led regulatory environment.
02:04Right.
02:04Because in order for us to progress, we have to have the guardrails.
02:09A regulatory environment that doesn't impede innovation but enables it.
02:15But with guardrails.
02:17And the second element to this is to ensure that we are not only attracting,
02:24but we are adapting.
02:26There is diffusion.
02:28And as we think about, if you look at the recent Microsoft report
02:32on AI adoption and diffusion, Malaysia is in the top 50.
02:36We have done reasonably well.
02:38Yeah, we've been ranking well and it's showing.
02:40It is ranking well.
02:42What we need to now look at is look at the entire tech stack.
02:46Everyone focuses on the shiny AI at the end,
02:50but there's an entire tech stack to it.
02:52And how we play a role in every layer of the tech stack.
02:57That would allow for sustained.
03:00Not only making sure that we are attracting, countries are attracting investment,
03:06but there is economic value at every layer.
03:09Exactly.
03:10So that would mean that we also need to be looking at what numbers we are doing for local procurement,
03:16ensuring that, of course, these data centers are also able to attract upskilled Malaysians to work.
03:23So it's not just about those big figures.
03:25Thanks, Dr. Jasmine.
03:25Pesa, now let's look at testing the investment-related case.
03:30Now, another big headline recently related to AI challenges was that S&P Global Ratings,
03:36they estimate Malaysian data center projects will require an additional 20 billion USD investment in funding over just the next
03:46three years.
03:46So that's by 2030.
03:48So some thoughts on that?
03:50Yeah, so I think that's, well, it's definitely in line with the kind of investment that's required, you know,
03:57to build, to maintain a data center.
03:58So definitely from my perspective, at least, I see that as a positive,
04:02because that also helps to deepen the capital markets in Malaysia, the financial markets here.
04:07I think you will see that a lot of the data centers that we have been, that's been coming out
04:11in Malaysia,
04:12a lot of them are funded both from financing, you will see that a lot in the press, a lot
04:16of announcements.
04:18That's quite common, I think, in the data center cycle, especially in the more, if you look at US, for
04:23example,
04:24in the Europe, it always starts first with bank loans, and then it goes on to perhaps to the bond
04:29market,
04:29and then to other more interesting financing structures, whether it's looking at a REIT,
04:34at specific data center REIT, specializing in one type of asset,
04:38or maybe even into securitization of that sort of a portfolio, which is a data center asset.
04:43So definitely, it's for that kind of financing that's required,
04:46I do see that that is going to be an evolution along the path from just bank financing all across
04:53the spectrum.
04:54Exactly. Now, another issue of interest would also be, Pace,
04:59that the fact that all of these billions still flowing in, billions still needed,
05:02they are also long-term assets, but they are being financed against assumptions,
05:08not just about AI demand, but also around the amount of natural resources that we have available,
05:14in particular, of course, water and then electricity, energy demands.
05:18So what should actually boards and investors be looking out for
05:23to ensure that what they are going to be projecting for the very near future
05:28is compatible with the resources available?
05:31Yeah. So I think that's a good question.
05:34So I think, interestingly, because for a lot of this data center investment,
05:38of course, most of it is looking at the asset itself,
05:41and more importantly, also who the offtaker is going to be,
05:43because ultimately, that is actually where the revenue stream is going to be coming from.
05:47So while, yes, there is a lot, I would say there are some uncertainties,
05:52but definitely that would have been priced in or factored in while at the onset,
05:57while they're looking to obtain their financing, whether it's from a debt perspective or whether from an equity perspective.
06:03A lot of the data centers that we see, there is also strong shareholder support as well,
06:09because I think they believe that in this sector that it continues to grow.
06:13So that combined with the external financing that they have,
06:17that allows that to have a good mix of both equity shareholder support as well as financing from other markets
06:23as well.
06:23Exactly. And at the same time, also be remaining cognizant that you need to give a little bit of leeway
06:29for higher power costs.
06:30And even the fact that almost every quarter we are seeing such advances in AI build,
06:36that infrastructure needs may not be what they appear to be today. Right.
06:42Dr. Jasmine, going into what China has been doing and China's launch of the World AI Cooperation Organization or WICO.
06:51Now, this appears to reflect growing competition over AI standardization, as well as infrastructure and digital partnerships.
06:59In your view, for countries like Malaysia, what kind of opportunities does this create for us,
07:05seeing that now we seem to have a more consolidated move towards not just AI sovereignty, but also partnerships intra
07:14-region?
07:15I think that's a very fair question.
07:17Malaysia is a member of ASEAN and ASEAN has remained neutral.
07:22We have adopted what makes sense to our industries across the region, not just Malaysia.
07:29The tech neutrality stand that Malaysia has taken is something I think everybody welcomes for different reasons.
07:37Different technology, different stacks have different use cases.
07:42And we recognize that.
07:45And I do want to call out the fact that you talked about AI sovereignty.
07:49It's interesting that the word sovereignty, there is no single definition.
07:55What, at the end of the day, governments want to have is control.
08:00They want to be able to say they know where the data is residing.
08:05And that comes in...
08:06And they always have access to it uninterrupted, naturally.
08:09That's right.
08:10Precisely.
08:11And that's where it is critical that we create regulations or develop and evolve our regulations and policies
08:20in a way that is relevant to current times and not tagged on to an analog environment, right?
08:27What used to work in analog will not work in digital because analog was very boxed and very siloed.
08:36The digital world interacts.
08:39Yeah.
08:39It speaks.
08:40I mean, if you let it, that is.
08:41Of course, right.
08:43So to me, I think when we think about usage and countries adapting, it is important that they understand what
08:53is their strength and what they want to build on.
08:57It is an AI economy is truly an innovation economy.
09:02It allows you to innovate.
09:04It allows the smallest company to innovate fastest because the barrier has been lowered.
09:13Access to technology, cost has been lowered.
09:16Exactly.
09:17So these are so many of the considerations out there.
09:20And together with that as well, Pejia, let's look at the financing, this growth, which is extremely positive at the
09:28moment,
09:28without mispricing the risk involved.
09:30Now, governments around the world are also really competing at the moment to attract AI investment.
09:36And this is through, they've got the whole stack tax incentives, you've got infrastructure spending and also appropriate policy support.
09:43At what point does competition for investment begin to reduce the value that countries are able to capture, though?
09:52Yeah.
09:52So I think that's something, I think you've seen quite a lot of it recently as well.
09:56I mean, the concern as to whether have we been rolling out the red carpet too much, right?
10:01Because, okay, I guess from my perspective, I think we should be looking at it holistically.
10:06So I totally agree that it's not just how much that investment size is going to be.
10:12But more importantly, what sort of value multiplier does that investment size create?
10:19And the value multiplier is not necessarily just in number and sense.
10:23Because I think as Malaysia moves towards, we are already there, right, at the precipice of moving into a high
10:30value industry.
10:31What we need to see, and from an investment perspective, to make sure that that balance is there,
10:36you know, when does it wear out the investment incentive versus the return that we get,
10:42is that we have to move away from just looking at the big number, but, you know, are we creating
10:47high-skilled jobs?
10:48And high-skilled jobs in which areas specifically?
10:52Because if we want to be truly in the AI economy, which is an innovation economy like Dr. Jasmine says,
10:57right,
10:57what kind of high-skilled jobs do we want these investors to really invest in,
11:02which then can have a knock-on effect to the Malaysians in this country, right?
11:07And when we talk about, as to be very specific, when we talk about technology transfer, what do we mean?
11:13What type of technology do we want to be transferred?
11:15We want high value, we want to have, you know, cutting-edge technology,
11:19not something that has been around and has just been amended, you know, or just updated.
11:24We want something that is truly brand new, something that can bring us to the next level.
11:28So I think those are the kind of stuff that we would need to look at in terms of ensuring
11:33that
11:34whatever incentive that's being given, whether it is the DSAC, which is the income tax incentive,
11:39or the customs duty exemption, also a lot of flexibility being granted to Malaysian digital companies
11:46in terms of getting financing as well, that has to be balanced with what we can see.
11:51So I think we are moving not just from employment, but what type of employment,
11:56moving away from technology transfer to what type of technology transfer.
12:00I think that is the next stage that we should be looking at.
12:02Yeah, the conversation and the economy also refines further.
12:06Thank you so far.
12:07We take a very quick break.
12:08Don't go anywhere.
12:09We'll be back with the rest of the conversation on Malaysia's AI economy challenges
12:12right here on Niaga Spotlight.
12:22Banyak 23 utas jam tangan mewah jenama Rolex dan 15 barang kemas yang disita SPRM
12:30daripada adik ipar bekas Panglima Tentera Darat secara rasmi menjadi harta kerajaan.
12:34Seramai 25,000 individu dijangka menerima manfaat melalui inisiatif Bakat Madani
12:39mulai pertengahan tahun 2026.
12:41Baga Zakat Selangor memberkasi asnaf menerusi projek agroekonomi laman agroehsan
12:47bagi melahirkan lebih ramai asnaf yang berdikari serta mempunyai sumber pendapatan yang mampan.
12:53It has been an exhibition from Marco Bersetti.
12:57Catch me if you can is the message.
13:00Game on in terms of the World Championship.
13:03We are back.
13:05MotoGP, we love you.
13:15Untuk semua pelanggan Astro, Chelsea menentang JDT.
13:199 Ogos, 8 malam secara langsung di Astro.
13:31Welcome back to Niagara Spotlight.
13:32Still with me, Tamina Kaosji.
13:34And the conversation continues on Malaysia's AI economy challenges
13:37with spokespersons from PwC Malaysia and Microsoft ASEAN right here in the studios with us.
13:43Dr. Jasmine, moving into the fact that AI, of course, is now not just a part of everyday working life,
13:48but also everyday business operations.
13:50But at the same time, adoption still largely depends on levels of trust,
13:56on also skills and, of course, clear governance guidelines.
14:00Looking across ASEAN, what are some valuable lessons which Malaysian operations can take from countries as well?
14:09Two things that we always look at, moving from pilot to productivity, right?
14:16We are cautious, which we understand, but we need to quickly move.
14:21Upscale, yeah.
14:22Yeah, and from pilot to productivity.
14:24And we need to get to production.
14:26We want to innovate quickly.
14:28We live in a generation that wants instantaneous gratification.
14:34And AI allows you to do that because it's able to crunch large amounts of data.
14:40And that brings us to why it is important for us to have structured data.
14:46We have a lot of data.
14:49We are data rich.
14:49But in order to have meaningful AI outputs, we need to have those insights.
14:55We are still not.
14:57We have insight poor, right?
14:59So we need to move to a culture.
15:03We need to embed.
15:05And this is our learning from across the world and across the region
15:09where countries have gone and moved into a very structured data
15:13where you can harness your insight and turn that into opportunity
15:17is where we see adoption.
15:20And that's where I think our businesses can start to think about
15:23how do they get their data structured, knowing their data estate
15:26and making sure they are turning that into insights, learnings.
15:30And as you know, product life cycles are becoming shorter
15:35because services and products now live shorter lifespan on shelves
15:40or digital shelves as it may be.
15:43And I think that's where our businesses can adapt and diffuse faster as well.
15:49And I think perhaps one thing to note is that, of course,
15:53the entire business sector, in particular SMEs,
15:56they have been very, very eager to adopt.
15:58But does purchasing more AI-related tools necessarily mean the ability to scale
16:05or would you caution against that?
16:08There is no one-size-fits-all, right?
16:11Depending on what kind of business,
16:13a startup would function very differently from a traditional business.
16:17And if you are a regulated industry,
16:20you have regulated the requirement of your regulated industry
16:24that you need to comply.
16:25Not only if you are a banking sector or you are in finance-related
16:29or healthcare solutions,
16:31then you have those guardrails that you need to comply with.
16:34And the lowest entry point is productivity.
16:40So how do you improve your business operations?
16:42But we want them to move up the value chain.
16:45We want them to use AI to innovate.
16:47So one has to know the baseline
16:51and look at where do you want to get to.
16:54I think that's the decision that companies need to make.
16:57Yeah.
16:58And I think something which strikes me
17:00as the visibilizing of many companies globally,
17:04not just in Malaysia or ASEAN,
17:05that may not yet have implemented that oversight,
17:09despite adopting,
17:10is the fact that quite often AI-related errors, etc.,
17:15they show up in produced output.
17:18So that can, of course, then visibilize
17:20that there isn't actually a structure to how it's being outputted.
17:24Because just because you can go so much faster,
17:26it doesn't mean you remove the human from the loop, right?
17:29I would always say this.
17:32You can delegate work to AI.
17:34You can't outsource accountability.
17:37Right?
17:38Yeah.
17:38Too true.
17:40Too true.
17:41Too true.
17:41So to me, that's an important reminder
17:44that we are outsourcing work,
17:46not outsourcing accountability.
17:48Absolutely.
17:49Thank you for the reminder.
17:50Pesa, let's move into climate resilience.
17:53Now, of course, data centers are expected to become
17:56one of our fastest-growing sources of electricity demand
18:00for the coming decade,
18:02combined with, of course, the strong investments.
18:04But in a region which is also facing rising temperatures
18:08as part of global warming overall,
18:10greater water stress,
18:12and also pressure on power infrastructure.
18:15What do you feel about climate resilience
18:18influencing investment decisions?
18:21And this ought to be something that is looked at by boards when?
18:26So I think it should be looked at right at the onset,
18:29if you ask me, right?
18:30Because I think that is a key part of ensuring
18:32that this continues to be sustainable.
18:34Because, you know, I like what Dr. Jasmine said, right?
18:36That, you know, we cannot outsource accountability.
18:38So I think that's very important.
18:40So how the data center,
18:42obviously there's a lot of benefits to that,
18:44but I think it's no secret
18:45that it's always going to be a cost.
18:47So how do we then ensure that cost
18:49continues to be sustainable?
18:51So that, to me, is something that boards,
18:53investors look at and to ensure that
18:56how that is being managed.
18:58So I think Malaysia, for example,
19:00for most of the data center investments,
19:02there's a task force that looks at it,
19:03the data center task force,
19:04which is, you know, a combined of all the bodies
19:07are led by Mida,
19:08where they actually assess from all angles,
19:10particularly what you're saying,
19:12particularly from a climate resilience perspective,
19:14you know, certain requirements that need to be met
19:16to ensure that it continues to meet
19:18what is globally accredited.
19:21I think there are also some data centers in Malaysia
19:22that have achieved accreditation
19:24that they have met a certain green index.
19:27So I think that is something
19:28that definitely would need to be looked at up front,
19:30in my view, not an afterthought,
19:32but as part of the entire process itself.
19:36Mm, exactly.
19:37I mean, just because there is going to be
19:40a notable uptick in energy intake,
19:43as well as electricity demand,
19:45that does not mean that we should continue
19:48to have the lack of visibility around these changes,
19:52which are also occurring at an intra-industry level.
19:55Perhaps it's also time then for those data centers,
19:58which have, for example, received their green accreditation
20:01to publicize this,
20:03because that also helps to allay
20:05the very well-founded fears of local communities
20:09when it comes to resource sharing, right?
20:11That's right, absolutely.
20:12And also, what would you say
20:14would actually help to distinguish
20:18genuine resilience planning
20:20from the opposite,
20:21which would be just a box-ticking sort of ESG exercise?
20:25So I think that, I mean, a lot of times
20:28is what you see in the outcome,
20:29because I think, you know, you can plan,
20:31but to ensure that it is genuine
20:33is really to see.
20:34And that's why constant monitoring is important
20:37to ensure that whatever that has been set out
20:41is definitely met as it continues to evolve,
20:44because it also depends on the changing needs
20:47or I would say the changing requirements,
20:49because as there are, you know,
20:52things continue to evolve,
20:53particularly from a climate perspective,
20:55if there's things that needs to be updated,
20:57I think that needs to be then modeled in
21:00or that needs to be considered
21:02during the monitoring process,
21:05because this is not something
21:06that is over one year.
21:07This is like a long-dated journey, right,
21:10that we are going.
21:11So that, to me, would need to track that against.
21:14Of course, there needs to be certainty
21:16and transparency in terms of the policies,
21:19because then that gives investors comfort.
21:22I don't think anyone comes in
21:24wanting to not do what is the right thing,
21:26but I think just wanting to know
21:27what is the right thing
21:29and how they can do what is required
21:32and even what is more than required,
21:34you know, how to do,
21:35how to make sure that that is being done well,
21:37as long as there is clear messaging
21:40and clear policies on how that is meant to be achieved.
21:43Yeah, and the goalposts can also, of course,
21:45shift with consideration to the evolving scenario.
21:49Thank you, Paisa.
21:50So as we look to landing the discussion,
21:51I've got a couple more quick ones for each of you.
21:54Dr. Jasmine, so much of the current conversation
21:57is just focusing around attracting infrastructure
21:59with regards to the AI economy,
22:01but Microsoft has experienced spanning cloud,
22:05AI, cybersecurity, as well as digital skills.
22:08Which capacity ought Malaysia to focus on
22:11within the next three to five years in particular?
22:15AI economy rests on three pillars.
22:19Infrastructure, trust, and talent, right?
22:24As we think about these three pillars,
22:27infrastructure is what we are attracting,
22:30the data centres.
22:31Trust is the governance piece.
22:33Yes.
22:33The certainty that Paisa talked about.
22:35It is critical that when we talk about trust,
22:40users, businesses will use technology that they trust,
22:44but they need to have that capability,
22:47and talent becomes a critical part of it.
22:51We have this ambition of becoming a high-income nation.
22:55We want to have the right GNI.
22:57We want to breach our GNI.
22:59We want to move away from a middle-income nation.
23:02And to become a high-income nation,
23:04we need to be able to create the right jobs.
23:07We can do all the skilling,
23:10but if we do not create the jobs,
23:12then it goes into this loop of either underemployment
23:16or unemployment.
23:19So, being able to provide that infrastructure,
23:23the trust, and the talent,
23:24because with trust and infrastructure,
23:26we would be able to attract the right kind of jobs, right?
23:31And then we create the talent.
23:33So, is it a talent first or a work first?
23:36It's a chicken.
23:37I will come in if you have talent,
23:39but if I don't come in, then I don't have jobs.
23:43So, it's a bit of a tug.
23:44It actually all has to be done simultaneously.
23:46Simultaneously.
23:47So, those three pillars.
23:48Thank you, Dr. Jasmine.
23:50Pesa, in conclusion, looking ahead,
23:53what would you say are some indicators
23:55Malaysia should be tracking over the next decade
23:58to judge whether all of this bounty of AI investments
24:03is actually going to strengthen the economy
24:05in a meaningful way?
24:06I think I'm looking forward to how we would need to track,
24:09again, is to see what kind of value
24:12that has created for the Malaysian economy,
24:15whether the kind of jobs that has been created.
24:19Also, I think it's also the overall multiplier effect.
24:23Of course, I think with all the influx of the data centers,
24:25actually, Malaysia has had data centers for a long time.
24:28But I think it's just really only come into
24:31a lot of the global presence now
24:33because, of course, of Johor,
24:35because of the hyperscalers.
24:36But we've actually had a long time
24:37even in Taman Technology, Bukit Jalil, in Cyberjaya.
24:40Exactly.
24:41It's just of a different size and scale.
24:42So, I think it has also pushed us to the forefront
24:45that, you know, I think if you look at a lot of global research,
24:48I think, especially in this part of the world,
24:50I think Johor itself, I mean,
24:51it's actually recognized as where most of the centers are booming.
24:55And Malaysia, for sure,
24:56because now it's also gone beyond Johor.
24:58It's also suddenly moving to other states as well.
25:00So, I think that also helps to put Malaysia on the global scene.
25:06It also attracts other sort of investment
25:08because data center is the infra, you know,
25:11it's not the end all as well.
25:13And AI is much larger than that, you know.
25:16It also attracts other players to come in as well
25:19to leverage on this value chain that we are creating
25:22because as the value chain develops to support the data centers,
25:27that value chain can also be adapted, modularized
25:30for other components of the AI economy.
25:33So, I think that is what we should be tracking.
25:35How are we growing as a nation together with, you know, the AI economy?
25:40Thank you, Paisa and Dr. Jasmine for the very insightful conversation.
25:43Well, that's all we have time for on today's conversation
25:47surrounding AI economy challenges.
25:49Thanks for watching Niagara Spotlight.
25:50We'll see you again next week with more business analysis and insights.
Comments