
This article first appeared in The Edge Malaysia Weekly on November 9, 2026 - November 15, 2026
MALAYSIA’s ambition to become an artificial intelligence (AI) nation by 2030 will see the government focus its resources on building the digital infrastructure while counting on private investment to drive the country’s transition from a consumer of technology to a creator of its own.
Digital Minister Gobind Singh Deo says Malaysia Digital 2030 (MD2030) — the latest phase of the digital transformation agenda — will not have a standalone budget. Instead, it will leverage existing government allocations, public-private partnerships and foreign and domestic private investment.
“We believe the strongest engine for achieving the scale and ambition of MD2030 will be private investment — both foreign direct investment (FDI) and domestic direct investment (DDI) — and public-private partnerships,” he tells The Edge in an email interview.
Public spending will focus on shared infrastructure, connectivity and digital platforms that lower costs for businesses developing new technologies.
“We want public investment to crowd in — not crowd out — private capital,” Gobind emphasises.
Incorporating the AI Action Plan 2026–2030, MD2030 seeks to raise the digital economy’s contribution to gross domestic product (GDP) to 30% by 2030, create 500,000 high-value digital jobs and spur the development of more Malaysian-made digital products, services and intellectual property (IP).
According to the Department of Statistics Malaysia (DOSM), information and communications technology (ICT) and e-commerce contributed RM476.8 billion or 23.5% of GDP in 2025, up from RM451.5 billion or 23.4% a year earlier.
However, translating that growth into domestic technological capabilities remains a challenge, particularly as Malaysia aims to move beyond hosting data centres and cloud infrastructure for global tech giants.
While foreign capital has expanded the country’s digital infrastructure, questions remain over how much it actually builds local AI capability.
Gobind notes that the government evaluates investments based on the economic value they generate, including tech transfer, opportunities for local suppliers and local IP development. “We are interested not simply in the size of an investment but in its economic value-add. Does it create high-value jobs? Develop Malaysian talent? Transfer technology and knowledge? Deepen local supply chains? Generate intellectual property? And strengthen the ability of Malaysian companies to compete globally?”
He stresses that the government does not regard FDI and DDI as competing sources of capital, saying both are needed to develop the domestic technology industry. Initiatives such as the Government-Linked Enterprises Activation and Reform Programme (GEAR-uP) are examples of how Malaysian capital can be mobilised towards strategic and high-value sectors, he says.
However, SNS Network Technology Bhd (KL:SNS) co-founder and managing director Ko Yun Hung observes that Malaysian businesses remain largely at the adoption stage while foreign companies account for much of the investment in AI infrastructure.
“At this time, most of it is still foreign. Local companies are more in the adoption phase; investing in AI infrastructure is not common yet,” he tells The Edge.
Ko believes domestic investment is more likely to grow as businesses become more comfortable with AI and identify commercial use cases. For instance, a company that initially subscribes to AI services to automate administrative work or analyse sales data may eventually find it worthwhile to invest in its own computing infrastructure as usage increases.
To accelerate this transition, Ko suggests policy incentives such as accelerated capital allowances for technology purchases and double tax deductions for AI-related workforce training. It may be more effective to introduce incentives and programmes to motivate adoption and usage than providing direct grants to every business, he says.
For most Malaysian businesses, the immediate hurdle is not developing sophisticated AI products but figuring out how to integrate the technology to streamline and improve their existing operations.
Ko notes that companies of various sizes are generally receptive to AI, though many remain uncertain about where to begin due to fragmented or poorly organised internal data.
While generative AI tools such as ChatGPT have made the technology more accessible, businesses are increasingly exploring customised applications that can perform specific tasks, such as agentic AI. Such systems can be configured to carry out more complex tasks with varying degrees of autonomy, such as managing a company’s human resources department’s leave applications and flag potential staffing shortages before managers make their decisions.
Yet, few local companies have progressed towards developing their own AI applications, much less products that can be sold commercially, Ko says. “Maybe not yet. In terms of moving from ChatGPT to something more like agentic AI, we are seeing that, but still not to the extent of developing their own applications.”
Cost is a key barrier, particularly for smaller companies that are unable to justify purchasing expensive graphics processing units (GPUs) before knowing whether their AI projects will deliver the intended results.
To address this, SNS launched its AI Factory last year, offering GPU-as-a-Servic (GPUaaS) to businesses seeking computing capacity for developing, testing or deploying AI applications without having to purchase the hardware themselves. Starting with 64 GPUs, the facility allows businesses to test and deploy applications without heavy capital expenditure.
Expanding access to such capabilities is central to MD2030’s ambition of developing more Malaysian “AI Builders” — individuals and businesses capable of creating applications, AI agents and commercially viable products and services, according to Gobind.
“We want more Malaysians who do not simply know how to use AI tools but who can build with AI — people who can develop applications, AI agents, products and services, solve real problems using AI and ultimately turn those capabilities into new businesses and solutions.”
The government plans to encourage industry-led training, intensive boot camps, apprenticeships and other non-traditional learning pathways, alongside university and technical education.
Gobind also acknowledges that retaining these talents requires creating sufficiently attractive opportunities in Malaysia, particularly for engineers, researchers and entrepreneurs who might otherwise pursue careers overseas.
For small and medium enterprises (SMEs), the emphasis is expected to shift from basic digitalisation, such as e-commerce and digital payments, towards using AI to improve productivity and eventually develop their own solutions.
Ko believes Malaysian SMEs can make that transition by 2030, but warns that the relatively short timeframe will require businesses and policymakers to move quickly. “I am confident that SMEs will be able to become builders. My only worry is the time. When you talk about the AI world, the competition and progress are at a very fast pace.”
Beyond investment and talent, Malaysia must consider how much control it needs over the technologies underpinning its AI economy, particularly when much of the global ecosystem depends on foreign-developed chips, cloud platforms and software.
Gobind says Malaysia should remain open to international technologies and partnerships while developing domestic capabilities in areas where it can compete. “The government should not prescribe where Malaysian companies must compete. The market will ultimately determine where Malaysia can build a competitive advantage.
“At the same time, there may be strategic areas where considerations of national interest, resilience and technological sovereignty require the government to provide greater direction or support.”
Farlina Said, director of the Cyber and Technology Policy programme at the Institute of Strategic and International Studies (ISIS) Malaysia, points out that technological sovereignty does not necessarily require countries to develop every component of their technology infrastructure domestically.
“Where one would view end-to-end indigenous production as true sovereignty, others may have to consider diversification and clear responsibilities,” she tells The Edge.
Malaysia nevertheless needs its own computing capabilities, she argues, as such infrastructure is critical to national security, economic development and the ability to respond to emerging technological risks, even if it remains difficult to determine how much domestic computing capacity Malaysia will need while the AI landscape is evolving.
In the meantime, the country should diversify its infrastructure, build locally relevant AI models and strengthen technical expertise to reduce vulnerabilities arising from dependence on foreign technologies.
She also calls attention to regulatory gaps involving the ownership and safety of AI training data, accountability when government agencies deploy AI and the responsibilities associated with autonomous AI systems.
Malaysia’s existing laws address areas such as personal data protection, cybersecurity and online safety, but Farlina says regulation alone will not be sufficient without the technical expertise to scrutinise increasingly sophisticated AI systems. “Much needed is to build the capability that pierces the black box, whether it is through internal technical knowledge or information sharing.”
With less than four years to 2030, execution across ministries will determine MD2030’s outcome.
“Malaysia has developed strong policies and blueprints over the years. The challenge has often been translating those ambitions into coordinated execution across government and the wider ecosystem,” Gobind admits.
MyDIGITAL Corporation will oversee programme management and monitor implementation while reporting directly to the National Digital Economy and Fourth Industrial Revolution Council, chaired by the prime minister.
“Ultimately, accountability must mean more than reporting whether a programme is on schedule. We need to know whether it is delivering the intended outcome — and, if it is not, what needs to change,” the minister says.
Farlina advises evaluating headline targets, including the digital economy’s 30% contribution to GDP and the creation of 500,000 high-value digital jobs, alongside broader socioeconomic outcomes, noting that employment targets may be harder to assess as AI itself disrupts existing jobs.
“I think Malaysia needs to see that the AI Nation serves national interests and we must find ways to illustrate whether it is on track or deviating from socioeconomic goals,” she says.
Ko agrees that investment metrics alone are insufficient to determine whether Malaysia has successfully developed a domestic AI industry. “I think the measurement should not just be about how much foreign or domestic investment we attract. A stronger measure is how much our SMEs are being enabled,” he says. “Because of the uniqueness of each business, they may develop something using AI tools for their own operations and later realise that what they have developed can also be offered as a service and sold to others.”
Save by subscribing to us for your print and/or digital copy.
P/S: The Edge is also available on Apple's App Store and Android's Google Play.