Companies in Southeast Asia are adopting artificial intelligence (AI) faster than the global average, but most are failing to turn the move into profits.
A survey of 330 companies across six Southeast Asian markets by McKinsey, Singapore’s Economic Development Board and Tech in Asia found that 46% have moved beyond pilots. However, most remain in the “experimentation trap”, or isolated AI projects that stop short of changing the workflows that drive profits.
The disconnect is evident in earnings, with 60% of companies reporting that AI adoption has had less than 5% impact on ebit. “It’s not about technology, but how do you really think about going for the adoptions and thinking about capturing the value,” says Vivek Lath, a partner at McKinsey Singapore who co-authored the study. “Companies are moving, but we are seeing only a small portion of the companies are able to capture the value.”
The performance divide centres on workflow. High performers (or those deriving significant earnings from AI) are twice as likely to fundamentally redesign how work gets done rather than simply layering AI onto existing processes. More than one in three allocate over 20% of their digital budgets to AI, several times the share of peers.
One executive told Lath that their company has more AI pilots than actual pilots in Singapore Airlines. "The challenge is that they are continuously thinking about small pilots or small use cases approach," he says.
Deeper change, he adds, calls for reimagining the whole operations, including redesigning workflows to embed AI, formalising governance and investing in the technology at scale and pace.
See also: China’s top chipmaker warns rushed AI capacity could sit idle
Execution gap
Despite strong enthusiasm, lack of internal AI expertise and talent was cited by 20% of respondents as the main barrier, followed by integration complexity at 16% and unclear return on investment at 12%. These obstacles indicate that execution, not vision, is the core issue in turning AI into value, asserts Lath.
He stresses that such execution depends on collective ownership. “AI is a team game, rather than an individual. The CEO should be the sponsor. The business leader will need to think about the problem. The HR person needs to think about how you’re building the skills. The risk person needs to think about what is allowed, what is not.”
See also: Jack Ma-backed Ant bets on AI health in US$69 bil sector race
The leadership focus must also shift from activity metrics to business impact. "From the beginning, think about how the solutions are going to be used by the frontline user, how are we going to ensure people adopt instead of fear that AI tool,” advises Lath.
Companies that got it right
A handful of regional groups show what that shift can deliver.
Petronas, Malaysia’s national oil and gas company, embedded AI directly into its corporate strategy. “The business strategy is our AI strategy. We wanted to enable the business strategy through AI or digital,” says Dr Rajamani Sambasivam, the company’s chief data scientist, in the report.
Doing so has enabled Petronas to deliver over 85% of its digital value from AI and data science initiatives. The company has upskilled over 26,000 employees and trained more than 5,000 to build machine-learning models to address operational problems.
Singapore’s DBS Bank followed a similar path after a decade of investment. “We aspire to be an AI-enabled bank with a heart,” says Nimish Panchmatia, its chief data and transformation officer, describing a shift from isolated use cases to changing how the bank interacts with customers and employees.
Capacity versus impact
To stay ahead of the latest tech trends, click here for DigitalEdge Section
Hyperscalers have invested more than US$50 billion in AI-ready data centres and cloud infrastructure across Southeast Asia, with billions more committed. This highlights the scale of the region’s technology bet.
Singapore has become a focal point of that AI buildout. Since 2024, more than 60 AI centres of excellence have been established in the city-state, according to the Singapore Economic Development Board. This spans technology firms such as Microsoft Research Asia and Google DeepMind as well as non-tech companies, including Prudential, Coca-Cola and Heineken, that are developing in-house capabilities to stay competitive.
But infrastructure alone will not guarantee returns. “Organisations in Southeast Asia are rapidly progressing on getting their infrastructure set up for scaling AI. However, there continues to be an opportunity to get the intelligence in terms of domains and use cases to get the return on investment from AI," says Saurish Basu, an associate partner at McKinsey Singapore who co-authored the study.
"The winners will be those who are able to take the opportunity to reimagine their business and workflows, rather than purely using AI to digitise existing processes," he adds.
