An International Labour Organisation study released this week paints a nuanced picture of artificial intelligence's emerging impact on the ASEAN workforce. While generative AI stands poised to influence the working lives of nearly 80 million people across the 11-nation bloc by 2025, current evidence suggests that large-scale employment losses remain absent, even as the technology reshapes occupational landscapes at an accelerating pace.

The comprehensive ILO report, titled "Generative AI and labour markets in ASEAN: Significant exposure, limited disruption, uneven preparedness," analysed both occupational exposure levels and real-world adoption patterns across the region. The findings reveal a complex reality that defies simple narratives of AI-driven unemployment. Instead, the data points to significant sectoral vulnerabilities coupled with substantial adaptation capacity in some economies, while others face institutional challenges in bridging emerging skills gaps.

According to ILO projections for 2025, approximately 22.9 per cent of total ASEAN employment—translating to nearly 80 million workers—occupies roles with more than minimal potential exposure to generative AI. Yet this headline figure masks a critical distinction that reshapes the interpretation of AI disruption risk. When the ILO narrowed its focus to occupations classified as facing the highest exposure category, the numbers contracted dramatically to just 3.3 per cent of the workforce, or approximately 11.7 million workers. Simultaneously, roughly two-thirds of ASEAN employment remains concentrated in occupations with no identifiable GenAI exposure whatsoever, suggesting that substantial portions of the labour market will continue functioning in traditional paradigms for years to come.

Geographic variation within ASEAN reveals telling disparities in technological vulnerability. Singapore emerges as the region's AI-exposed outlier, with 42.2 per cent of its workforce engaged in roles vulnerable to generative AI—a reflection of its advanced, finance and technology-driven economy. The Philippines follows with 28.1 per cent exposure, attributable partly to its robust service sector and growing information technology industry. Indonesia, the region's most populous economy, records 21.7 per cent exposure, while Viet Nam and Thailand trail at 20.8 per cent and 20.6 per cent respectively. These divergences underscore how economic structure, sectoral composition, and developmental stage determine AI vulnerability curves across Southeast Asia.

Designated as high-exposure occupations continue expanding throughout ASEAN despite the absence of visible mass displacement. This counterintuitive trend suggests that GenAI adoption is occurring alongside rather than instead of traditional job creation. The ILO report explicitly notes that "the potential for labour market transformation is significant, but widespread disruption is not yet visible," a finding that inverts fears of imminent technological unemployment. This characterization reflects the technology's current early-stage deployment across the region, where implementation remains concentrated within technology-intensive sectors rather than diffusing broadly through administrative and office-based roles, despite those roles theoretically facing considerable exposure.

Demographic fault lines emerge clearly when examining AI exposure patterns by age and gender. Young workers aged 15 to 24 experience exposure levels comparable to their adult counterparts, suggesting that the coming AI transition will not disproportionately disadvantage new labour market entrants relative to established workers. However, a pronounced gender dimension complicates this picture. Women demonstrate more than twice the occupational exposure to GenAI compared to men, primarily because they concentrate disproportionately in clerical, administrative, and professional service roles that generative AI tools can readily automate or transform. This gendered vulnerability emerges as perhaps the most acute labour market risk requiring targeted policy intervention.

The preparedness landscape across ASEAN presents an even starker picture of regional inequality. Singapore stands singularly positioned as a globally competitive AI ecosystem, combining cutting-edge digital infrastructure, abundant high-skill talent, and comprehensive governmental coordination across multiple ministries and agencies. The city-state's whole-of-government implementation strategy contrasts sharply with less-developed peers, where fragmented approaches, skills shortages, and limited access to capital constrain AI adoption and upskilling capacity. This preparedness gap risks widening over time, potentially concentrating AI-era productivity gains and employment opportunities among regional leaders while leaving others vulnerable to disruption without accompanying transition support.

For Malaysian policymakers and business leaders, the ILO findings carry specific implications. Malaysia's diversified economy—spanning manufacturing, services, and emerging technology sectors—likely experiences exposure levels somewhere between Singapore and Indonesia, though comprehensive national-level data remain limited. The country's established role in regional business process outsourcing and growing fintech presence mean that administrative, customer service, and professional roles face meaningful AI displacement risks. Simultaneously, Malaysia's manufacturing base and vast informal economy retain substantial segments insulated from near-term technological disruption.

The report outlines critical priorities for minimising disruption while capturing AI's productivity potential. These include establishing human-centred governance frameworks that position worker welfare alongside technological progress; dramatically expanding upskilling and reskilling programmes with explicit attention to women and youth populations; supporting micro, small and medium enterprises in overcoming adoption barriers; and strengthening knowledge exchange mechanisms across ASEAN member states. Malaysia, positioned as a relatively advanced regional economy with developed institutional capacity, could assume leadership in facilitating these cross-border initiatives.

Medium-term labour market transformation appears inevitable, but its trajectory and distributional consequences remain malleable through deliberate policy intervention. The absence of large-scale job losses to date should not encourage complacency, as technological adoption curves often accelerate exponentially. Rather, this window of limited disruption represents a critical opportunity for ASEAN governments and businesses to proactively develop skills ecosystems, strengthen social protection systems, and ensure that AI-driven productivity gains translate into broadly shared prosperity rather than concentrated advantage.

The ILO's findings ultimately suggest that the ASEAN AI transition mirrors previous technological revolutions in following unpredictable paths that combine genuine disruption in some sectors with unexpected expansion in others. Singapore's advanced positioning, the Philippines' service sector concentration, and Indonesia's demographic scale present distinct adaptation challenges and opportunities. Malaysia's intermediate position and institutional capacity suggest capability to navigate these transitions successfully, provided that policymakers act decisively during this relatively calm period before AI adoption accelerates substantially across the region's labour markets.