CASE STUDY
February 2026
Executive Summary
Singapore occupies a paradoxical position in the global AI transition: it is simultaneously one of the world’s most AI-ready economies and among the most exposed to AI-driven labour displacement. With 730 industrial robots per 10,000 employees — the second-highest robot density in the world — and over 75% of industries already integrating AI into workflows, the structural transformation of Singapore’s labour market is well underway.
This case study identifies the principal categories of workers at risk, disaggregates vulnerability by demographic profile, evaluates the economic outlook under current trajectories, and assesses the adequacy of existing and proposed policy interventions. The central argument is that while Singapore’s institutional infrastructure for workforce adaptation is among the most developed globally, structural gaps — particularly concerning younger workers, women in administrative roles, and migrant labour — require targeted redress if the benefits of AI adoption are to be broadly shared.
- The Landscape of AI Exposure in Singapore
1.1 Scale and Pace of Adoption
According to the Infocomm Media Development Authority (IMDA), more than 75% of industries worldwide have integrated AI into their workflows as of 2024, a figure reflected in Singapore’s own corporate landscape. Research by Workday found that 79% of organisations in Singapore are already deploying AI agents or have begun operating them, signalling a shift from pilot programmes to full-scale operational integration. Singapore’s robot density has grown at an average annual rate of 27% since 2015.
The 2026 National Budget, delivered by Prime Minister Lawrence Wong, acknowledged this reality directly. Calling ‘jobs, jobs, jobs’ the government’s number one priority, the Budget established a National AI Council — chaired by the Prime Minister — to coordinate AI deployment across four strategic sectors: advanced manufacturing, transport and connectivity, finance, and healthcare. The formation of this body reflects recognition that AI adoption has moved beyond voluntary corporate experimentation and now demands active macro-coordination.
1.2 Structural Exposure of the Workforce
An IMF Selected Issues Paper (Khan, 2024) provides the most granular available assessment of Singapore’s workforce exposure. The paper estimates that Singapore’s labour force is ‘highly exposed’ to AI technologies, owing to its large share of skilled, white-collar workers — precisely the segment most susceptible to generative AI substitution. Of this highly exposed segment:
Approximately 38.9% are employed in occupations with high AI complementarity — including managers, science and engineering professionals, health professionals, and legal professionals. These workers stand to gain productivity benefits from AI augmentation.
Approximately 38.6% are in occupations with low AI complementarity — including clerical support workers, business and administration professionals, ICT associates, and a portion of sales workers. These individuals face heightened substitution risk.
The symmetry of this split — roughly half benefiting, half at risk — masks important distributional inequalities, detailed in the following section.
- At-Risk Worker Profiles
2.1 Occupational Categories at Greatest Risk
The following table summarises the primary occupational categories facing high AI exposure and low complementarity in Singapore’s context:
Occupation Category AI Exposure Adaptability Key Risk Factors
Clerical & Administrative Support High Low High task automation potential; narrow skill transferability
Business & Administration Professionals High Moderate Routine decision-making and reporting tasks most exposed
ICT Associates / Entry-Level Tech High Moderate Code generation AI compressing entry-level demand
Sales & Customer Service High Low Chatbots and AI agents replacing front-line interaction
Contract / Gig Workers (incl. DBS bank roles) High Low No retraining support; first to be released in cost reductions
Low-Wage Migrant Workers (manufacturing/logistics) Moderate-High Very Low Excluded from SkillsFuture; high robotisation in their sectors
A particularly concrete indicator of displacement pressure is DBS Bank’s announcement that it expects to cut approximately 4,000 contract roles over three years as AI increasingly automates these functions. Contract workers — who lack the employment protections of permanent staff and are frequently ineligible for government retraining subsidies — represent a structurally exposed category.
2.2 Demographic Concentrations of Risk
The IMF paper provides a critical demographic dimension to the occupational data. Two groups warrant particular analytical attention:
Women: Women are disproportionately concentrated in clerical, administrative, and sales occupations — the precise roles identified as having high AI exposure and low complementarity. This mirrors findings from the Brookings Institution’s parallel research on U.S. workers, where 86% of the high-risk, low-adaptability cohort were found to be women. In Singapore’s context, the IMF estimates that female workers face systematically greater vulnerability to the disruptive effects of AI, raising concerns about widening gender-based income inequality in the absence of targeted policy intervention.
Younger Workers: Counter-intuitively, younger Singaporean workers — often assumed to be more digitally capable — face elevated risk. An estimated half of workers aged under 30 are employed in occupations with high AI exposure and low complementarity, compared to only one-fifth who are in high-exposure but high-complementarity roles. Younger workers are disproportionately represented in clerical and sales positions, which carry limited AI complementarity, and are underrepresented in the managerial and senior professional roles that stand to gain from AI augmentation.
Migrant Workers: Singapore hosts a substantial migrant workforce concentrated in construction, domestic services, manufacturing, and logistics. These workers are largely excluded from SkillsFuture and related national upskilling programmes. As the CSIS notes, many already carry substantial debt from basic certification requirements, making self-financed reskilling financially inaccessible. The combination of sectoral robotisation pressure and programme exclusion renders this population particularly precarious.
- Economic Outlook
3.1 Near-Term Displacement Risks
The Civil Service College’s scenario analysis describes a plausible near-term trajectory in which ‘by 2025, knowledge workers across domains have been significantly displaced due to AI augmentation’ — with professionals in law, brokerage, and management consulting facing pay cuts or layoffs, and ‘job redesign’ becoming a euphemism for reduced compensation. Consumer spending contraction and mortgage stress are identified as downstream economic consequences. While this represents one scenario rather than a forecast, its underlying logic is consistent with the structural data.
The pace of AI adoption, rather than its eventual scale, is the primary risk variable. Singapore’s existing safety nets — unemployment assistance, retraining grants, job matching services — were calibrated for more gradual labour market transitions. Rapid AI-driven displacement could exceed the absorption capacity of these systems, particularly if multiple sectors are disrupted in parallel.
3.2 Medium-Term Structural Effects
The World Economic Forum’s Future of Jobs Report 2025 projects that AI and information-processing technologies will create 170 million new roles globally by 2030 while displacing 92 million, yielding a net addition of approximately 78 million jobs — 14% of current global employment. For Singapore, the UNDP projects that AI could lift annual GDP growth by approximately two percentage points and raise productivity by up to 5% in sectors such as health and finance.
However, the distributional character of these gains is critical. The ICT sector’s share of total job postings has nearly halved over the past twelve years in Singapore, even as absolute job numbers in the sector have grown. This compression of entry-level demand — particularly in technology — has significant implications for graduate employment and wages at the lower end of the skill distribution.
The skills mismatch risk is compounding. The WEF estimates that 39% of key skills required in the labour market will change by 2030, down slightly from 44% in 2023 — suggesting some stabilisation in the rate of skill obsolescence, but still representing a substantial retraining burden across the entire working population.
3.3 The Inequality Vector
Perhaps the most significant economic risk is not aggregate displacement but distributional divergence. The bifurcation of the workforce into AI-complemented high earners and AI-substituted low earners, if left unaddressed by redistribution and retraining policy, could materially worsen Singapore’s Gini coefficient. The IMF paper is explicit: in the absence of appropriate policies, AI adoption ‘could worsen income inequality in Singapore.’ The demographic concentration of risk among women and younger workers intersects with pre-existing structural inequalities, potentially compounding them.
- Policy Responses: Assessment and Gaps
4.1 Existing Frameworks
Singapore’s policy architecture for workforce transition is comparatively well-developed. Key instruments include:
SkillsFuture Initiative (est. 2015): Provides citizens and permanent residents over 25 with subsidised access to thousands of courses and career transition support. The SkillsFuture Level-Up Programme (SFLP), recently enhanced, offers mid-career workers aged 40+ up to S$3,000/month in full-time training allowances, with a new S$300/month part-time allowance from early 2026.
Budget 2026 Measures: Six months of free access to premium AI tools for Singaporeans enrolled in selected courses; productivity support grants for firms; National AI Council formation for strategic sector coordination; redesigned SkillsFuture digital platform for improved accessibility.
Employment Pass and Job Support Schemes: Ongoing labour market interventions targeting involuntary unemployment, with planned expansion of temporary income support.
Model AI Governance Framework for Agentic AI: Establishes guardrails for AI deployment with human accountability requirements, designed to ensure responsible adoption rather than unchecked automation.
4.2 Critical Gaps
Despite this infrastructure, several significant gaps warrant attention:
Exclusion of Migrant Workers: The most structurally vulnerable group in Singapore’s labour force — low-wage migrant workers — is almost entirely excluded from national reskilling programmes. This is both a policy failure and an ethical concern, particularly given Singapore’s stated commitment to inclusive workforce development.
Youth-Oriented Programming Deficit: Current SkillsFuture programming predominantly targets mid-career individuals aged 40 and above. The IMF paper specifically recommends expanding these initiatives to include younger workers, who face disproportionate high-exposure, low-complementarity employment profiles.
SME Adoption Barriers: Large corporations such as DBS, PSA, and multinational technology firms can invest in sophisticated AI implementation and workforce transition simultaneously. Smaller enterprises — which constitute the majority of Singapore’s business landscape — face cost, complexity, and change management barriers that may impede both AI adoption and worker transition support.
Skills Curriculum Obsolescence Risk: The rapid pace of AI capability development risks rendering recently completed training obsolete. SkillsFuture’s provider network must continuously update curricula, but the quality and speed of this updating process across a large and heterogeneous provider base remains a governance challenge.
Private Sector Engagement: Singapore’s 2026 Budget commentary from the business community acknowledges that government-led reskilling ‘will be for naught if businesses and investors do not follow.’ The responsibility for sustained employability increasingly rests with firms, whose incentives to retrain workers rather than replace them are not guaranteed.
- Recommended Solutions
5.1 Extend SkillsFuture Access to Migrant Workers
A phased extension of SkillsFuture access to long-term work permit holders, beginning with those in sectors facing the highest robotisation risk, would address the most acute gap in the current framework. This could be structured as a graduated subsidy — covering a meaningful share of course costs without creating dependency — and coordinated with bilateral agreements with workers’ home countries in ASEAN, consistent with Singapore’s regional AI leadership ambitions.
5.2 Establish an Early-Career AI Transition Programme
A dedicated programme targeting workers under 30 in high-exposure, low-complementarity roles would fill the demographic gap in existing provision. This should emphasise practical AI tool use, with competencies aligned to near-term employer demand rather than theoretical digital literacy. The free AI tool access announced in Budget 2026 provides a useful foundation, but should be structured around supervised, workplace-relevant learning rather than self-directed consumption.
5.3 Introduce Sectoral Transition Compacts
For sectors facing particularly rapid displacement — including financial services, administrative services, and certain ICT roles — government, industry associations, and unions should negotiate forward-looking transition compacts: binding commitments by large employers to retrain rather than simply release affected workers, in exchange for productivity support and regulatory facilitation of AI deployment. This mechanism exists in various forms in European labour markets and could be adapted to Singapore’s tripartite industrial relations framework.
5.4 Strengthen the Temporary Unemployment Safety Net
The planned introduction of temporary income support for involuntary unemployment is timely but must be designed with AI-specific displacement pathways in mind. This includes ensuring that workers displaced by AI — who may not meet traditional redundancy criteria if their contracts are simply not renewed — are eligible for support, and that benefit duration is calibrated to realistic retraining timescales rather than historical job-matching periods.
5.5 Mandate AI Impact Assessments for Large Employers
Requiring large employers above a specified headcount threshold to conduct and publish annual AI workforce impact assessments — identifying roles at risk, retraining investments made, and transition outcomes for affected workers — would generate the data infrastructure currently lacking for evidence-based policymaking. It would also create accountability pressure that supplements voluntary corporate commitments.
- Conclusion
Singapore’s AI transition presents one of the most instructive case studies globally, precisely because the country combines advanced AI adoption capacity with a sophisticated policy infrastructure and a genuine commitment to social inclusion. The challenge is not capability but calibration: ensuring that the speed and distributional character of AI-driven labour market change do not outpace the adaptive mechanisms the government has constructed.
The 6.1 million workers identified in the Brookings Institution’s global analysis as occupying the ‘high exposure, low adaptability’ quadrant have their Singaporean counterparts in the clerical, administrative, sales, and migrant labour segments of the workforce. For these workers, the risks are real and the timelines are compressed. The existing policy framework provides a strong foundation, but the gaps — in youth-oriented programming, migrant worker inclusion, private sector accountability, and safety net design — require deliberate and urgent attention.
Singapore’s Prime Minister has stated that ‘fear cannot be Singapore’s response’ to AI. What is required instead is precision: targeted, evidence-based intervention that matches the specificity of AI’s distributional effects with equally specific policy remedies. The analytical tools to identify affected workers exist. The institutional capacity to act exists. The question is whether the political will to close the remaining gaps will materialise at the pace the labour market requires.
References and Key Sources
International Monetary Fund (2024). Impact of AI on Singapore’s Labour Market. Selected Issues Papers 2024, No. 040. IMF eLibrary. https://doi.org/10.5089/9798400285721.018
Brookings Institution (2026). AI Workforce Exposure and Adaptability Study. Research led by Sam Manning, Senior Research Fellow.
Civil Service College Singapore (2024). AI, Technology & Singapore: Preparing for the Future. Centre for Strategic Futures. https://knowledge.csc.gov.sg
CSIS New Perspectives on Asia (2021). Artificial Intelligence and the Future of Singapore’s Foreign Workforce. Centre for Strategic and International Studies.
UNDP Asia-Pacific (2025). Millions of Jobs at Risk in Asia-Pacific as AI Adoption Surges. UN News, December 2025.
World Economic Forum (2025). Future of Jobs Report 2025. WEF, Geneva.
HRM Asia (2026). Singapore Budget 2026: Scaling AI Through National Coordination and Workforce Integration. February 2026.
Eco-Business (2026). Singapore’s Response to AI Risks Will Fail Unless Businesses and Investors Follow. February 2026.
Noy, S. and Zhang, W. (2023). Experimental Evidence on the Productivity Effects of Generative Artificial Intelligence. Science, 381(6654).