The Future of Work Through an Economic Lens: Productivity, Labour Market Adjustment, and Human Capital in the Age of Generative AI

Over the last month, the future of work has moved from a broad strategic theme to an urgent economic question. Recent publications and events from the IMF, OECD, ILO, and World Bank show a clear pattern: labour markets remain relatively resilient in many economies, but the foundations of work are shifting because of generative AI, changing skill demand, uneven digital adoption, and rising uncertainty in the global economy. The central issue is no longer whether work will change, but how quickly productivity gains can be translated into better jobs, stronger firms, and more inclusive growth. This article examines the future of work through an economic lens. It argues that the next phase of labour market change will be shaped by five interacting forces: technological diffusion, productivity effects, wage polarization, demographic and educational adjustment, and public policy capacity. The article also shows that the economic effects of AI are unlikely to be uniform. High-income economies may capture earlier gains through stronger digital infrastructure and deeper capital markets, while many developing economies face the risk of disruption before receiving large productivity benefits. At the same time, recent evidence suggests that women, young workers, and routine office occupations may be more exposed to task transformation than others. The economic future of work, therefore, is not simply a story about automation. It is a story about institutions, incentives, training systems, labour market design, and the ability of societies to manage transition fairly. For higher education institutions such as Swiss International University (SIU), this debate is especially relevant, because the value of education will increasingly depend on whether learners can combine analytical knowledge, adaptive skills, and ethical judgement in a labour market shaped by intelligent technologies.

Keywords:Future of work; generative AI; productivity; labour economics; human capital; skills; inequality; digital transformation; higher education; economic policy

Introduction

During the last month, the global discussion around work has intensified in a more concrete and economic direction. The OECD’s international conference on AI in work, innovation, productivity and skills, held from 30 March to 1 April 2026, highlighted how fast the policy conversation has moved toward agentic AI, algorithmic management, skills, and productivity. Around the same time, the ILO and World Bank drew attention to the uneven global effects of generative AI, while the IMF’s April 2026 World Economic Outlook placed AI-related productivity expectations alongside war, fragmentation, and renewed trade tensions as major factors shaping the world economy. This combination is important. It shows that the future of work is no longer a narrow technology topic. It has become part of mainstream macroeconomic analysis.

This matters because labour markets are currently sending mixed signals. On one hand, OECD unemployment remained at 5.0 percent in February 2026, still low by historical standards, and employment rates were stable in many member countries in late 2025. On the other hand, job vacancies have eased in several advanced economies, youth unemployment has risen faster than overall unemployment, and many firms are rethinking hiring patterns as digital tools improve. In economic terms, this is a classic adjustment period: the quantity of jobs may not collapse, but the composition of jobs, the returns to skills, and the structure of productivity are changing underneath the surface.

A useful economic perspective begins with a simple distinction. Technology affects labour markets in at least three ways. First, it can substitute for labour in particular tasks. Second, it can augment labour by making workers more productive. Third, it can create entirely new tasks, occupations, and business models. The overall economic result depends on the balance between these effects, the speed of adoption, the institutional environment, and the distribution of gains. For this reason, the future of work should not be discussed in purely technological language. It should be studied as a question of productivity, labour demand, wage formation, and social adjustment.

The current trend is especially significant because evidence is becoming more precise. OECD material notes that generative AI can improve performance in specific tasks by about 20 to 40 percent under favourable conditions, while the broader economy-wide effects remain uncertain. At the same time, IMF analysis points to wage premiums for new skills, but also suggests that demand for AI-related skills has not yet translated into broad employment growth in the same way as other new skills. In some regions, employment levels in AI-vulnerable occupations were lower where demand for AI skills was higher, especially after several years. This suggests that firms may be rewarding specialized digital capability while reducing recruitment into more automatable entry-level roles. That is a very different pattern from traditional stories of innovation-led job expansion.

The rest of this article argues that the future of work through an economic lens can be understood through five themes: productivity and diffusion; labour market polarization; the geography of unequal gains; human capital and education; and the policy challenge of managing a fair transition.

1. Productivity: The Core Economic Promise

The strongest economic argument in favor of AI is still productivity. Across long periods, living standards depend heavily on productivity growth. When output per worker rises, firms can produce more with the same inputs, wages can grow more sustainably, and governments gain more fiscal space. This is why economists pay so much attention to new general-purpose technologies. In principle, AI has the features of such a technology: it can be applied across sectors, improve over time, and affect many business processes at once. OECD analysis notes that generative AI tools can significantly improve performance in particular tasks, and IMF commentary similarly frames AI as a potential engine of global growth.

Yet the important phrase is under favourable conditions. Productivity does not rise automatically when a new tool is introduced. Historically, large gains appear only when firms redesign workflows, invest in complementary skills, update management practices, and spread technology widely rather than keeping it inside a few frontier firms. This matters today because the real economic prize is not simply faster text generation or automated summaries. The prize is organizational transformation: better forecasting, improved customer response, faster research, more accurate administration, lower transaction costs, and new service models. If adoption remains shallow, the macroeconomic effect will disappoint. The IMF’s April 2026 outlook explicitly includes disappointment over AI-driven productivity as a downside risk to growth.

Productivity also has a distributional side. If AI mainly helps already-productive firms, already-educated workers, and already-digitized sectors, the average productivity number may improve while inequality deepens. This is a common pattern in digital economies: frontier companies capture large gains first, while smaller firms lag because of cost, weak data systems, low managerial capacity, or limited digital skills. The result can be a widening gap between leading and lagging enterprises. In economic terms, diffusion becomes the key variable. A country may have powerful AI tools available in the market, but unless ordinary firms can absorb them, national productivity effects will remain concentrated and incomplete.

This is why the future of work debate cannot be separated from management quality. Management is the bridge between technology and outcomes. Two firms may have access to the same tools and obtain very different results. One uses AI to reduce low-value repetitive work and free employees for more complex tasks. Another uses it mainly as a cost-cutting instrument and weakens trust, morale, and learning. The first firm is more likely to convert digital capability into lasting productivity. The second may gain briefly but lose longer-term adaptability. Therefore, from an economic perspective, the future of work is not only about machines becoming more capable. It is also about whether organizations learn to combine technology with human judgement, coordination, and professional accountability.

2. Labour Market Adjustment: Why Job Loss and Job Creation Will Coexist

Public debate often asks whether AI will destroy jobs. Economically, that question is too simple. It is more accurate to ask which tasks will be automated, which jobs will be re-designed, which occupations will expand, and who will bear the adjustment costs. The ILO’s recent work continues to show that exposure to generative AI is widespread but uneven. Earlier ILO findings indicated that one in four workers globally are in occupations with some exposure to generative AI, while only a smaller share falls into the highest exposure category. More recent ILO publications in March and April 2026 sharpen this point by emphasizing the unequal structure of exposure across countries and social groups.

This means that many occupations will not disappear suddenly, but many tasks inside occupations may change quickly. Administrative support, routine office work, scheduling, documentation, customer communication, and some analytical drafting tasks are already more exposed than occupations requiring complex manual, interpersonal, or context-heavy judgement. The likely result is not a single wave of mass unemployment across all sectors. Instead, labour markets may experience task reallocation. Workers who once spent much of their time on routine cognitive tasks may be expected to supervise systems, check outputs, manage exceptions, interpret results, and communicate with clients or teams. That shift can increase productivity, but it can also create stress for workers who are not trained for the new task mix.

Economic theory suggests that when technology substitutes for tasks in the middle of the skill distribution while complementing high-skill analytical roles and preserving some lower-skill service roles, labour markets become polarized. Recent IMF commentary is consistent with this concern, noting that high-skill and low-skill workers tend to gain the most while middle-skill roles, especially routine office jobs, are being squeezed. This is important for social mobility. In many economies, middle-skill occupations have long provided stable entry points into the middle class. If those roles shrink or become thinner stepping stones, young workers may face a more difficult path into secure employment.

There is also evidence that entry-level hiring may be especially vulnerable. IMF analysis notes that employment levels in AI-vulnerable occupations are lower in regions with higher demand for AI skills, and it cites emerging evidence that generative AI adoption reduces entry-level hiring where tasks are more automatable. This is a serious economic issue because labour markets do not only allocate income; they also create experience. If entry-level positions shrink too quickly, economies may save labour costs today but weaken their future talent pipeline. The result could be a paradoxical labour market in which employers complain about skill shortages while providing fewer opportunities for people to gain those skills through work.

This is one reason why the future of work must be approached as a transition problem. Even if long-run innovation creates new occupations, the short- and medium-term adjustment can still be painful. Workers do not move instantly from declining tasks into growing sectors. They need retraining, information, time, and often income support. Economists call these frictions adjustment costs. They are central, not secondary, to the future of work debate. A technologically advanced economy can still produce weak social outcomes if it ignores the transition costs faced by real workers and families.

3. Unequal Gains Across Countries: The Risk of “Disruption Without Dividend”

One of the most important developments in the last month is the stronger attention to global inequality within the AI debate. The new ILO and World Bank work warns that many developing countries may experience disruption before receiving major benefits. The phrase “disruption without dividend” is economically powerful because it captures a real risk: if countries lack digital infrastructure, advanced firms, reliable electricity, data systems, financing, and training capacity, then workers may face changing labour demand without seeing large productivity growth, wage gains, or new formal employment opportunities.

This does not mean developing economies are protected from AI. In fact, exposure may arrive through trade, outsourcing, business process change, and global value chains even when domestic adoption is weak. A company in one country may automate support functions and reduce demand for outsourced labour elsewhere. A platform may change service models globally. International clients may require new digital standards from suppliers. Therefore, weaker digital ecosystems do not guarantee safety. They may simply reduce local capacity to benefit from change.

Recent ILO and World Bank materials also note that the structure of exposure differs by income level. In high-income countries, a larger share of jobs is exposed to generative AI, because more workers are employed in cognitive and office-based occupations. In low-income countries, exposure is lower on average, but the augmentation potential is also weaker if adoption remains limited. World Bank material released this month similarly notes that job exposure to AI automation is lower in developing countries than in advanced economies, but that this does not automatically mean better outcomes, because lower adoption can limit productivity gains. This is why the economic question is not only “Who is exposed?” but also “Who has the institutional capacity to convert exposure into growth?”

For universities and business schools, this global dimension matters deeply. The future of work in Europe, the Gulf, Africa, Asia, and Latin America will not follow one identical path. Labour market institutions differ. Informality differs. Access to technology differs. Industrial structure differs. In some regions, the priority may be advanced AI governance and professional upskilling. In others, the more urgent need may be digital basics, entrepreneurship support, and pathways from informal to formal employment. A serious academic discussion must therefore avoid a single universal model. The economics of the future of work will be shaped by local constraints as much as by global technology.

4. Gender, Youth, and the Social Distribution of Risk

A high-level economic analysis must also examine who is most exposed. Recent ILO publications released in March 2026 show that women face higher workplace risks from generative AI than men in most countries studied, largely because women are more concentrated in clerical and administrative tasks that are more automatable. Search summaries from ILO sources report that women are more exposed in 88 percent of countries in the sample, and in high-income countries overall, 41 percent of jobs are exposed to generative AI compared with 11 percent in low-income countries.

This does not imply that women will simply lose more jobs in a mechanical way. But it does mean that gender inequality could be reshaped through task transformation, wage pressure, and promotion pathways. If administrative and coordination work becomes more automated, some traditional occupational ladders may weaken. At the same time, if women have unequal access to advanced digital training, leadership roles, or AI design functions, a new gender gap could appear within the digital economy itself. The issue is therefore not only exposure, but also mobility: who moves into augmented, supervisory, analytical, and strategic roles as workplaces change.

Youth face a different but related challenge. OECD data show that the unemployment rate for younger workers in OECD countries rose to 11.5 percent in February 2026, much higher than for older groups. IMF analysis adds that entry-level jobs have higher exposure to AI. Together these points suggest an economic risk: the first stage of a career may become less stable at the exact moment when workers most need experience, confidence, and income. If the bottom rung of the ladder becomes thinner, inequality can widen over time even if headline employment looks stable.

From a policy perspective, this means that the future of work is inseparable from the economics of opportunity. Productivity growth is valuable, but if access to good jobs narrows, societies may become richer on average while less open in practice. The quality of transition matters as much as the speed of innovation. Countries that invest in targeted reskilling, fair recruitment, and broad digital access are more likely to achieve inclusive gains. Countries that assume the market will solve these issues by itself may face deeper inequality and lower social trust.

5. Human Capital, Universities, and the New Value of Education

The future of work is often discussed as a labour market story, but it is equally a human capital story. If tasks change faster, then the economic value of education depends less on static knowledge alone and more on adaptive capability. Workers will need to learn, unlearn, and relearn across longer careers. This makes the role of universities more important, not less. However, it also changes what universities must deliver.

The strongest educational response is unlikely to be narrow technical training by itself. Technical skills matter, but the future labour market also rewards judgement, problem framing, communication, ethics, interdisciplinary reasoning, and the ability to evaluate machine outputs critically. OECD material on AI and education stresses both the promise of personalized learning and the need to prevent digital divides, bias, and poor safeguards. In the labour market, this means graduates will need both digital fluency and intellectual maturity. They must know how to work with intelligent systems, but also how to challenge them.

For Swiss International University (SIU), this debate is directly relevant. A modern international university cannot prepare students only for the jobs of the present. It must prepare them for economies where occupations evolve, work is reorganized around data and digital tools, and professional value increasingly comes from combining knowledge with adaptability. In that context, higher education should strengthen at least four capacities: analytical thinking, technological literacy, ethical reasoning, and continuous learning. These capacities are economically valuable because they support labour mobility. A worker with such skills is better able to move across sectors, absorb new tools, and contribute in roles that are not easily reduced to routine execution. This is the real long-term protection against technological volatility.

At the same time, education systems must not become detached from real labour market needs. Employers increasingly value people who can interpret information, collaborate across functions, use digital systems effectively, and make sound decisions under uncertainty. The future of work therefore requires a closer relationship between academic learning and applied capability. This does not mean reducing education to short-term market demand. It means recognizing that rigorous knowledge and employability are not opposites. In a volatile economy, deep knowledge is useful precisely because it supports adaptation.

6. Policy, Institutions, and the Economics of a Fair Transition

The final economic lesson is that the future of work will be shaped by institutions. Markets alone can generate innovation, but they do not automatically produce fair transitions. If productivity rises while labour protections weaken, bargaining power declines, or access to reskilling remains unequal, then the gains from technology will be concentrated. That is why recent OECD, ILO, and IMF materials all emphasize policy, skills, and social protection rather than technology alone.

A fair transition requires several things. First, countries need credible lifelong learning systems. One-time education at the beginning of adulthood is no longer enough. Second, labour market data and forecasting must improve so workers and institutions can identify changing demand early. Third, social protection systems must support people during transitions rather than waiting until displacement becomes severe. Fourth, governance matters. Algorithmic management, data use, and AI deployment in workplaces should be transparent enough to protect workers’ dignity, privacy, and procedural fairness. The recent OECD conference agenda itself reflects this shift, placing algorithmic management and policy response at the center of the discussion.

There is also a macroeconomic dimension. The IMF’s April 2026 outlook reminds us that the future of work is unfolding during geopolitical conflict, public debt pressure, inflation risks, and fragile growth. This means governments cannot rely on ideal conditions. They must support adjustment while managing fiscal constraints. In practice, this increases the importance of smart prioritization: invest in broad digital infrastructure, targeted education, competitive markets, and institutions that help firms adopt technology productively rather than merely using it to cut costs. The long-run goal should be not just more AI, but better economic transformation.

Conclusion

The future of work through an economic lens is not a simple narrative of machines replacing people. It is a more demanding and more interesting story. Recent evidence from the last month shows that labour markets remain relatively resilient, yet they are entering a period of deeper structural change. AI may raise productivity, but only if firms, workers, and institutions can absorb it effectively. Jobs may not disappear all at once, but tasks, career paths, and skill premiums are already changing. Some workers and countries may gain quickly, while others face disruption without immediate reward. Women and young workers may face particular risks. Developing economies may need different strategies from advanced ones. And universities will play a major role in determining whether societies produce adaptable, thoughtful, and capable graduates for this new environment.

Economically, the central challenge is distribution. The question is not only whether AI can create value, but who captures that value, how widely it spreads, and whether the transition is managed in a socially sustainable way. The future of work will reward societies that combine innovation with education, productivity with fairness, and technological ambition with institutional wisdom. In that sense, the future of work is not only about the economy of tomorrow. It is also about the policy and educational choices made today.

Hashtags:

#FutureOfWork #GenerativeAI #LabourEconomics #Productivity #HumanCapital

Sources

1. International Monetary Fund, World Economic Outlook, April 2026: Global Economy in the Shadow of War

2. OECD, Labour Market Situation, Updated April 2026

3. OECD, 2026 International Conference on AI in Work, Innovation, Productivity and Skills

4. OECD, Artificial Intelligence topic page, updated 2026

5. International Monetary Fund Blog, New Skills and AI Are Reshaping the Future of Work

6. International Labour Organization, Disruption without Dividend? How the Digital Divide and Task Differences Split GenAI’s Global Impact

7. International Labour Organization, Gen AI, Occupational Segregation and Gender Equality in the World of Work

8. ILO–World Bank paper on the uneven global impact of generative AI on jobs

9. World Bank, World Development Report 2026 Concept Note

10. OECD, The Impact of Artificial Intelligence on Productivity, Distribution and Growth