South-east Asia battles with talent crunch and ecosystem gaps in global AI race

THE rapid adoption of artificial intelligence (AI) across South-east Asia has sparked an intense race to tap the region’s talent pool, but a limited supply of skilled practitioners is creating significant challenges for businesses looking to harness this transformative technology.

Lyon Poh, technology consulting partner at KPMG, told The Business Times that he believes the supply of skilled technology and AI talents to be limited globally, with the shortage especially among individuals who can both envision and execute complex projects.

An analysis by the professional-services firm found that an estimated 30 to 70 per cent of data, security and development job postings are going unfilled. By 2025, the global shortfall of full-time software developers and cybersecurity professionals is expected to hit eight million.

Yeo Puay Lim, commercial director of Singapore-headquartered recruitment service platform Glints, said the biggest challenge to AI adoption “remains the shortage of skilled AI practitioners, including software engineers, data engineers and researchers”.

He added: “Despite the proliferation of readily available foundational models today and the ramp-up in educational infrastructure, the skills to create, implement, deploy and evaluate AI systems in a production capacity are still sorely needed.”

The scarcity has pushed startups to undertake cross-border hiring – that is, tap the regional talent pool across South-east Asia, rather than just scout for talent locally.

Yeo explained: “The emergence of young, skilled, technically competent AI talent from improving Stem (science, technology, engineering and mathematics) education, a growing digital economy, and a startup ecosystem in countries like Indonesia and Vietnam in the past 10 years have built a wider talent pool for startups to access, while managing their use of capital.”

Focus on usage

An estimated 40 to 45 per cent of South-east Asian companies now use AI at scale, up from the 15 per cent reported in a 2020 survey by consulting firm Kearney, which polled more than 100 companies. Glints found that among 72 startups surveyed in Singapore, Indonesia, Vietnam and Taiwan, 43 per cent were planning to explore or implement AI to manage costs and reduce overheads within the next 12 months. But Yeo expressed less concern about the race for talent between startups and Big Tech companies, because startups are more likely to focus on AI usage than on developing foundational models, which has high barriers to entry.

Big Tech vs startups

Dr Wu Pei Chuan, associate professor in the Department of Management and Organisation at NUS Business School, agreed that AI talents can be grouped into two distinct categories: technical experts who can build AI systems, and domain business experts who can effectively employ AI in their respective fields, such as finance and healthcare.

She added: “Although Big Tech companies have well-developed infrastructure and many challenging projects, startups can be attractive to certain groups of people with their innovative edge.”

KPMG’s Poh further pointed out that promoting from within and investing in current employees can reduce hiring costs and increase employee engagement, benefiting smaller businesses competing for talent with large tech companies.

KPMG’s recent Skills Data project with FTSE 100 organisations, which are the top 100 companies listed on the London Stock Exchange by market capitalisation, found that 79 per cent of open roles can be filled by upskilling existing talent (54 per cent) or redeployment (25 per cent).

Poh noted: “To compete with Big Tech for talent, businesses need to be more creative by offering outcome-based rewards and purposeful career pathways within their industries.”

Generic applications

Industry watchers agree that a fragmented AI ecosystem poses another challenge for AI adoption in the region.

Varun Arora, a partner at Kearney, observed that despite the recent surge in AI use cases, customer-interaction applications remain generic and underdeveloped. He highlighted two main issues – poor data quality and the need for local language adaptations in most South-east Asian countries. Additionally, many companies are hindered by outdated tech systems, limiting their ability to implement advanced functionalities, he added.

Glints’ Yeo echoed this sentiment. “Each country in the region has unique languages, cultures and regulatory environments. The availability of accurate data for foundational models also varies. A model built for one location could underperform or be irrelevant in another.”

Ensuring compliance when using data and AI models across markets is another challenge for businesses. Singapore has been driving growth in this area through the Digital Economy Partnership Agreement with Chile and New Zealand, signed in 2020.

Arora envisions that a common data residency framework would enable the region to reap the benefits of AI, as individual South-east Asian countries may not be significant markets on their own. Creating an ecosystem that targets all markets synergistically could be more financially viable for startups. “This approach enables the sharing of infrastructure, such as data centres, by allowing some segments of data to reside outside the country,” he noted. “It also facilitates the sharing of the AI ecosystem across countries. For example, startups can work on data from multiple countries to create applications, train them on multi-country, multi-language data, and deploy them at a regional level.”

Carving out niches

Industry watchers say that the United States remains the undisputed leader in the AI race, but agree that South-east Asia can carve out its niche with context-specific innovations and strategic initiatives.

Indonesia is a key player to watch in the partnership space, Arora said. “While there have been recent announcements about Big Tech ramping up investments in Malaysia and Thailand, Indonesia already received one round of investments two years ago.”

Glints’ Yeo added that governments in the region are also playing a pivotal role by implementing forward-thinking policies and investing in AI research and development. Besides Singapore’s National AI Strategy, which aims to transform key sectors of the economy using AI by 2030, Indonesia, Vietnam and Malaysia have each formulated their national frameworks around AI.

Yeo highlighted that South-east Asia’s strength lies in its youthful, tech-savvy population. The region is witnessing a rapid growth in its digital economy and startup ecosystem, and companies are using AI to address specific challenges in healthcare, agriculture, finance and logistics.

He added that this trend has been particularly prominent in Singapore, Indonesia and Vietnam.