Policy

23% of Southeast Asian Workers Face AI Job Disruption

White-collar roles show highest exposure while manual labor remains largely insulated, ILO analysis finds.

Omega Editorial· August 13, 2026· 2 min read

White-collar jobs face greater AI exposure than manual labor

Nearly one in four workers across Southeast Asia will see their jobs disrupted or significantly altered by generative AI, according to research from the International Labour Organization published in July. The analysis challenges assumptions that automation primarily threatens manual labor, instead identifying office-based roles as most vulnerable to transformation.

Clerical, administrative and professional positions face the highest exposure to generative AI technologies. Manual trades, craft work and agricultural jobs remain relatively protected from AI-driven changes, the ILO found. The degree of impact varies significantly across the region depending on factors including economic development levels, digital infrastructure maturity and existing labor policies.

Why it matters

Business leaders planning workforce strategies need to understand that AI exposure doesn't equal job elimination. The research suggests a more complex transformation where roles evolve rather than disappear, requiring investment in reskilling programs and workforce planning that accounts for changing skill requirements rather than simple headcount reduction.

Most workers remain unaffected

The data reveals a more nuanced picture than early automation forecasts suggested. Only 3.3 percent of Southeast Asian workers hold positions in the highest exposure category, while approximately 67 percent work in roles with no identified generative AI exposure whatsoever.

"Our understanding of automation and its impact on jobs has become more nuanced than in the early forecasts surrounding the Fourth Industrial Revolution," said Phu Huynh, senior employment specialist at the ILO.

Craft workers, trades professionals and skilled agricultural workers show "generally low" exposure to generative AI disruption, according to Huynh.

Exposure means evolution, not elimination

Analysts emphasize that high AI exposure doesn't translate to wholesale job replacement. Instead, affected roles are likely to shift toward higher-value activities or evolve to incorporate AI tools as productivity enhancers.

This distinction matters for both employers and workers. Organizations face decisions about retraining existing staff versus hiring for new skill sets, while workers in exposed categories should consider how their roles might transform rather than assuming displacement.

The regional variation in impact points to the importance of local context. Countries with more advanced digital infrastructure and supportive labor policies may see smoother transitions, while those with less developed technological foundations could face sharper disruptions.

These findings were first reported by the South China Morning Post, drawing on the International Labour Organization's July analysis of generative AI's workforce impact across Southeast Asia.

#generative ai#workforce transformation#southeast asia#international labour organization#automation#future of work

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in Policy

Policy· 3 min read

Twitch Introduces AI Training Opt-Out After Years of Data Use

Amazon's streaming platform confirms user content has fed generative AI models since at least 2024, with creators automatically enrolled.

Via AI Watch · Aug 12, 2026
Policy· 3 min read

AI Agents' Deceptive Behavior Erodes User Trust, Report Finds

Autonomous systems that lie and manipulate are creating a credibility crisis that demands new governance frameworks.

Via AI Watch · Aug 12, 2026
Policy· 3 min read

White House Plans to Expand AI Oversight Framework

Trump administration officials are expected to revise current AI guidelines to include open-source models, according to sources familiar with the discussions.

Via WIRED · Aug 12, 2026