The International Labour Organisation has released findings indicating that generative artificial intelligence will have far-reaching consequences for the ASEAN workforce, with exposure spanning nearly 80 million workers across the 11-nation bloc. However, the research offers reassurance that widespread displacement remains unlikely in the near term, even as the technology reshapes labour dynamics across the region.
According to ILO projections for 2025, approximately 22.9 per cent of total employment in ASEAN faces more than minimal exposure to generative AI. This proportion, while substantial, masks significant variation in the intensity of potential disruption. Only 3.3 per cent of workers—roughly 11.7 million people—operate in occupations classified as facing the highest level of AI exposure, suggesting that most affected workers will experience manageable rather than transformative change. The remaining two-thirds of employment remains in occupations with no identified exposure risk, providing a buffer against wholesale labour market upheaval.
The exposure levels vary dramatically across individual ASEAN economies, revealing stark differences in how integrated each nation has become with digital and service sectors. Singapore leads significantly with 42.2 per cent of its workforce facing more than minimal AI exposure, reflecting its advanced digital economy and technology-focused industries. The Philippines follows with 28.1 per cent, a figure that underscores how substantially the country has built its service and IT-oriented sectors over recent decades. Indonesia, the region's most populous nation, records 21.7 per cent exposure, while Vietnam and Thailand cluster at 20.8 per cent and 20.6 per cent respectively. This ranking directly correlates with each economy's existing technological infrastructure and sectoral composition.
A particularly striking finding concerns gender disparities in AI exposure. Women face disproportionate vulnerability, with female workers more than twice as likely as men to work in occupations classified as having high GenAI exposure. This concentration reflects the historical clustering of women in clerical, administrative, and professional roles—precisely the functions most amenable to AI automation. The disparity signals potential labour market inequality unless policymakers implement targeted interventions, and represents a critical consideration for Malaysia, where female workforce participation has grown substantially over the past decade.
Youth face comparable exposure levels to older workers, suggesting that age alone does not determine vulnerability to AI disruption. This finding challenges conventional assumptions about generational AI adoption and indicates that young people entering the workforce will encounter fundamentally different career trajectories than their predecessors, regardless of their relative comfort with new technology. The implications extend beyond individual workers to educational institutions, which must recalibrate training programmes to reflect these realities.
Currently, adoption of generative AI tools remains uneven and concentrated within technology-intensive sectors. Despite high theoretical exposure, office and administrative roles—where exposure should theoretically drive rapid adoption—show comparatively limited uptake of AI solutions. This gap between exposure and actual implementation suggests that technological capability alone does not drive automation; organisational readiness, regulatory frameworks, cost considerations, and workforce resistance all influence real-world deployment patterns. For Malaysian employers, this indicates a transition period during which skills redundancy may increase gradually rather than abruptly.
The report emphasises that while labour market transformation potential looms large, visible disruption has not yet materialised across the region. Employment in highly exposed occupations has continued to expand, contradicting predictions of imminent contraction in vulnerable sectors. This pattern suggests that companies are augmenting rather than replacing human workers with AI tools, at least in the initial adoption phases. Such dynamics provide a window of opportunity for policymakers and employers to prepare workforces for transition.
Regional preparedness for this transformation remains decidedly uneven. Singapore emerges as exceptional, possessing an integrated approach combining advanced digital infrastructure, abundant AI talent, and coordinated governmental implementation strategy. Most other ASEAN members lag significantly behind, creating a two-tier region where developed digital economies capture AI productivity gains while less-advanced economies struggle to harness the technology effectively. Malaysia occupies a middle position, with established IT sectors in cities like Kuala Lumpur and Selangor, yet still facing challenges in universal digital literacy and infrastructure across all states.
The ILO has identified several regional priorities to guide AI transition management. Human-centred governance frameworks must balance innovation incentives with worker protections, ensuring that technological progress benefits employees rather than displacing them without support. Inclusive skills development programmes require expansion of upskilling and reskilling opportunities, with particular emphasis on reaching women and youth—the groups most vulnerable to disruption or least equipped to navigate it. Micro, small and medium enterprises, which collectively employ vast numbers of ASEAN workers, require targeted assistance to overcome barriers to AI adoption, preventing a scenario where only large corporations benefit from efficiency gains.
Knowledge exchange and coordinated human resource development across ASEAN member states could help prevent a race to the bottom in labour standards as companies seek competitive advantage through cost reduction. Regional coordination mechanisms might enable smaller economies to share Singapore's lessons and infrastructure developments rather than duplicating expensive pilot programmes independently. For Malaysia specifically, initiatives that build upon existing IT competence while extending preparedness to less-developed regions could position the country as a regional training and innovation hub.
The immediate challenge facing policymakers involves bridging the preparedness gap before AI adoption accelerates beyond the current uneven, technology-sector-concentrated phase. Training institutions must shift curricula toward skills that complement rather than compete with AI, emphasising creativity, emotional intelligence, and complex problem-solving. Simultaneously, social safety nets require strengthening to support workers displaced during transition periods, particularly in sectors with limited alternative employment opportunities. Malaysia's existing technical and vocational training frameworks, whilst established, may require substantial enhancement to meet projected demand for reskilling programmes.
The window for proactive intervention remains open but narrow. The coincidence of low visible disruption and expanding employment in exposed sectors suggests that the transition from current labour market structures to AI-integrated economies will unfold gradually over years rather than months. This timeline provides sufficient opportunity for comprehensive policy development, institutional reform, and workforce preparation—but only if governments, employers, and educational institutions act decisively now. The difference between ASEAN economies that manage this transition successfully and those that stumble will likely determine their economic competitiveness for decades to come.
