The Real Economics of AI and Jobs

AI is fundamentally transforming the global job market, driving profound changes in skill requirements, entire professions, and wage structure across both advanced and emerging economies. Predictions about AI’s labor market impact range drastically: from mass worker displacement, to a productivity revival, to somewhere in the middle. 

The World Economic Forum is optimistic that job creation will outpace job losses in the near term. However, this labor market transformation will be complex and challenging. But navigating the transition will require not only an understanding of technological innovation but also coordinated efforts in policy, education, and workforce development.

[time-brightcove not-tgx=”true”]

Our ability to successfully adapt will depend on the speed of AI integration. Steam power, textile machinery, electricity, internal combustion engines, and personal computers all took 20 to 40 years for a widespread labor market impact to materialize. The internet, containers, spreadsheets, automated telephone switching, and mechanized agriculture all advanced faster, over 10 to 20 years. 

The integration of AI across economies is expected to be faster still—generating an investment boom that may become a bubble. And yet, the timeline of its labor market impact is less clear-cut. In my discussions with leaders and experts, some believe that the pace of integration will be so rapid that vast swaths of the labor market will be displaced, and a new “rustbelt” will emerge in some of today’s white-collar hotspots, from Manhattan and London to Bangalore and Dubai. Others are more sanguine: AI integration will be more gradual and will augment rather than displace workers, allowing time for workforces, governments, and employers to adjusts and thrive. 

What these predictions miss is the relative sophistication and complexity of our economies. We need to look beyond how one specific technology might replace specific tasks and instead focus on building resilience that can adapt to a wide range of technological and global shifts. 

Geoeconomic trends could drive the trajectory of labor markets in equal, if not greater, measure than technological change. As trade and foreign direct investment fall in employment-intensive sectors like infrastructure and traditional manufacturing—where each direct job typically creates 2.2 indirect jobs—the future of globalization-enabled jobs is uncertain. 

In the UK, government estimates indicate that the number of jobs created through foreign direct investment has reduced by 3% and that the country is experiencing its lowest number of such projects since records began 18 years ago. In Ciudad Juárez, Mexico, which borders the United States, reduced trade due to tariff uncertainty led to an estimated 64,000 factory job losses between 2023 and 2025. 

At the same time, the new multipolar, competitive order is also creating new geoeconomic booms and with them, wholly new job opportunities: from defence industries in South Korea, Turkey, and Poland, and chipmaking in Malaysia; to critical minerals in Australia and food and agriculture in Brazil. In these regions, a new set of job opportunities will draw in talent and create multiplier effects in adjacent sectors.

Demographics will also impact jobs over the next few years, including how quickly AI-based labor substitution occurs. As immigration barriers in many advanced economies rise, concurrently with ageing and talent shortages, the propensity to automate tasks will increase. Early-stage developments on this front are already evident in Japan, where historically tight immigration controls and the world’s most advanced aging society have pushed the country to trial innovative automation methods such as eldercare robotics.

The interplay between technology and demographics in developing economies will be more uncertain. Over the next decade, an unprecedented 1.2 billion young people in developing economies will enter the workforce. On the one hand, with abundant human talent, the political pressure for domestic job creation will be high. On the other hand, if labor-displacing technology is cheap enough, it will rapidly reduce traditional job opportunities for young people. Historically, there are examples of both–comparative advantage on the basis of lower, skilled cheap labor such as in the early stages of China’s export manufacturing boom in the 1980s, or Bangladesh’s textile industry in the 90s, as well as higher-skilled adoption of new technologies such as in higher value add manufacturing in East Asia or the IT industry in India through the 1990s and beyond. 

In this more multifaceted outlook, how can policymakers, employers, and workers plan for what lies ahead?

In all economies, both high- and low-income, one clear win-win move for policymakers and businesses alike is to drive rapid change in lifelong learning systems.

Adapting lifelong learning need not further draw down on stretched fiscal capacity or constrained budgets. Instead, it demands innovation in how this funding is deployed: modernizing public job centers and career guidance infrastructure, upgrading job data to cutting-edge, real-time labor market and skills information systems, and collaboration between universities, business, and government to deliver skills at scale.  And we must also adapt our education systems to ensure we are providing students with skills for the economy of tomorrow, including AI, digital, human-centric, business, green, and vocational skills. The Nordic economies have long been leaders on this front. So has Singapore, with its SkillsFuture initiative, and Brazil, where a skills accelerator seeks to link upskilling to job market demand while embedding digital skills into education programs. 

Even in the AI boom itself, it has become clear that the commercial viability of AI investments remains a distant prospect without a commensurate investment in the AI literacy skills executives and staff need to generate new, creative uses of the technology across fields, from healthcare and education to agriculture and finance. Evidence from recent studies about healthcare AI adoption shows that even when AI tools and infrastructure are widely deployed, for example, in radiology across dozens of U.S. healthcare systems, meaningful clinical impact often lags due to insufficient training, workflow integration challenges, and clinician readiness.

For many developing economies, the way forward will lie in combining vast talent and relatively cheap technology to create the equivalent of an “industrial policy” for entrepreneurship. A strategic approach to financing and supporting entrepreneurship—including freelancing, small businesses, and digital enterprises—especially in high-demand sectors such as software development, digital marketing, consulting, and creative services, can create win-win opportunities for youth while driving domestic growth. Some countries are already moving on this front; for example, Nigeria’s National Talent Export Programme aims to position the country as an outsourcing hub by aligning local companies, development partners, and government. The alternative is a generation of youth in the global south facing diminishing prospects for social mobility and growing societal strife. 

Even though AI promises to transform our economy, a narrow focus on any one technology, including AI, risks driving the wrong conclusions and investments when it comes to jobs. Policymakers, business, and workers alike need to consider demographics, geoeconomics, and technology—to design the right talent strategy for today’s economy.