AI Layoff Wave Hasn’t Hit ASEAN, But the Labour Market Alarm Bells Are Ringing

An ILO brief finds that nearly 80 million workers in ASEAN are employed in occupations with some degree of exposure to generative AI, but only a small share are in the highest-exposure jobs. The findings matter because the region’s challenge is not yet mass displacement, but uneven preparedness, gendered exposure, youth entry-level pressures and the risk that productivity gains concentrate in better-equipped firms and economies.

Generative AI is already touching a sizeable share of work across ASEAN, but the clearest risk is not immediate mass unemployment. It is a more uneven transition in which women, young workers, smaller firms and less-prepared economies may face the sharpest adjustment pressures.

According to an ILO brief, nearly 80 million workers across ASEAN are employed in occupations with some degree of potential exposure to generative AI, equivalent to 22.9 per cent of total employment. Yet the highest-risk category remains much smaller: only 3.3 per cent of total employment, or about 11.7 million workers, is concentrated in occupations with the highest levels of exposure. Around 67 per cent of employment remains in occupations with no identified GenAI exposure.

So far, the regional evidence does not support a simple "AI is taking jobs" narrative. Employment in ASEAN occupations with at least minimal GenAI exposure rose from around 66 million workers in 2017 to 74 million in 2022 and 80 million in 2025. The growth should not be credited to AI itself. It also reflects broader shifts toward services, knowledge work, structural transformation and labour-force growth, but it does show that the most exposed occupations have not yet seen a region-wide employment collapse.

Women and young workers sit closer to the fault line

The ILO brief, titled 'Generative AI and labour markets in ASEAN: Significant exposure, limited disruption, uneven preparedness', reveals that women are significantly more likely than men to work in the occupations most exposed to GenAI. Across ASEAN, 4.8 per cent of women were employed in highly exposed occupations in 2025, compared with 2.3 per cent of men. The gap reflects women's concentration in clerical, administrative and selected professional occupations, while men remain more concentrated in manual and operational work with lower exposure.

It does not mean women will automatically lose from AI. Some exposed roles are skilled and relatively well-paid, but it implies that AI policy cannot be gender-blind. If women are concentrated in roles where tasks are more exposed, while remaining underrepresented in STEM and AI-related occupations, the transition could widen existing labour-market inequalities unless reskilling, career mobility and workplace protections are built into policy responses.

Young workers present a different warning signal. Youth and adult exposure levels are broadly comparable across ASEAN, but in some countries young workers are slightly more likely to be in highly exposed jobs. Indonesia and the Philippines stand out, with 4.0 per cent and 4.6 per cent of youth employment respectively concentrated in the highest exposure category, compared with 3.2 per cent and 3.9 per cent among adults.

More concerning, youth employment in highly exposed occupations declined markedly in the Philippines and Thailand relative to total youth employment, while Viet Nam's trends remained more aligned. The available evidence does not show a corresponding increase in youth unemployment or youth not in employment, education or training.

The youth signal may reflect changing hiring patterns, education choices or occupational shifts rather than direct AI displacement. Still, entry-level roles are often where employers first redesign tasks, raise skill expectations or test automation. For ASEAN's young workers, the risk may not be that jobs vanish overnight, but that the first rung of the career ladder becomes harder to reach.

The real divide is between adoption and preparedness

AI use is rising across ASEAN, but not evenly. The share of people using GenAI products increased across all ASEAN countries between the first and second half of 2025, with particularly strong uptake in Singapore and Viet Nam. Yet usage in several countries remains below the global average, and the data captures general use rather than workplace adoption alone.

GenAI use remains concentrated in technology-intensive and knowledge-based occupations. Available usage data shows stronger activity in computer and mathematical roles and education-related tasks, while office and administrative support occupations show more limited observed use despite being prominent among highly exposed occupational groups.

Firm size adds another divide. In Singapore, a 2026 establishment survey found that 71.5 per cent of surveyed firms had not begun AI adoption, while only 3.8 per cent had integrated AI into core business processes. Large firms were much more likely to adopt AI than smaller firms, and adoption was strongest in information and communication, professional services, and financial and insurance services.

This is where the economic stakes become clearer. If AI adoption remains concentrated among larger firms, digitally advanced sectors and better-prepared economies, productivity gains may accrue to those already best positioned to benefit. Smaller enterprises, informal workers and less-connected labour markets could face competitive pressure without receiving the same tools, training or institutional support.

ASEAN's next AI battle is policy, not prediction

ASEAN's AI future will not be determined by exposure numbers alone. It will depend on whether governments, firms and labour institutions can convert technological change into productivity gains without deepening inequality. The region's preparedness is uneven. Singapore combines high exposure with high readiness, while other economies show different mixes of exposure, digital capacity and institutional strength.

A three-tier landscape is emerging: Singapore as a globally competitive AI ecosystem; Malaysia, Thailand, Brunei Darussalam, Indonesia, the Philippines and Viet Nam with important foundations but continuing gaps; and Cambodia, Lao PDR, Myanmar and Timor-Leste with lower readiness linked to more limited digital infrastructure, institutional capacity and technological capability.

Regional AI governance is also still largely soft-law and coordination-oriented. ASEAN frameworks promote human-centred and innovation-friendly AI, but implementation remains mostly national. Labour and social dimensions are recognised, including employment disruption and social protection risks, yet governance remains focused more on guidance, cooperation and multi-stakeholder engagement than binding labour standards or formal tripartite mechanisms.

The transition will require digital infrastructure, enterprise capability, affordable connectivity, labour-market intelligence, social protection, gender-responsive training and stronger worker-employer dialogue.