12 Sep The AI Talent Paradox: Companies Are Looking Globally for Skills They Must Also Learn to Build
Companies facing talent shortages have an increasingly large world in which to look for workers. But one of the biggest workforce challenges emerging in 2026 may not be solved simply by searching farther.
Artificial intelligence is changing the skills employers need faster than labor markets can readily supply them. At the same time, AI is beginning to alter the entry-level jobs through which companies traditionally develop the experienced workers they will need later.
That creates a paradox: employers are looking around the world for scarce skills while potentially disrupting the very talent pipelines that create those skills in the first place.
Research from ManpowerGroup, Atlas HXM and the World Economic Forum suggests that solving the problem will require companies to think about talent acquisition and talent development as parts of the same strategy.
AI has changed what is difficult to hire
ManpowerGroup’s 2026 Talent Shortage Survey provides perhaps the clearest indication of how quickly employer demand is changing.
Its survey of more than 39,000 employers across 41 countries found that 72% are having difficulty filling positions. For the first time, AI capabilities have overtaken traditional engineering and IT skills as the hardest skills for employers to find.
AI Model & Application Development was cited by 20% of employers and AI Literacy by 19%, placing them at the top of the shortage rankings. Yet the research also shows that demand is hardly limited to technical knowledge. Communication, collaboration and teamwork were the most sought-after human attributes, cited by 39% of employers, followed by professionalism and work ethic at 36% and adaptability and willingness to learn at 34%.
That combination is significant. Employers are not simply looking for people who know how to operate new technology. They increasingly need people who can use it with judgment, collaborate with others and continue learning as the technology changes.
Companies are responding accordingly. Upskilling and reskilling is the most commonly cited employer response, while others are increasing schedule and location flexibility, raising wages and targeting new talent pools.
When talent is scarce, the search becomes global
One consequence is that employers have greater reason to look beyond their traditional labor markets.
A Metaintro analysis of Atlas HXM’s Global Atlas Report 2026 found that 49% of organizations with international workforces describe attracting and retaining global talent as very or extremely challenging. Yet companies are not responding by retreating from international recruitment.
Among U.S. organizations surveyed, 67% said changing immigration policies were actually accelerating workforce and hiring decisions. Employers may consider different locations, remote arrangements, relocation or other forms of cross-border employment when a particular talent market becomes difficult to access.
This turns global hiring into more than an expansion strategy. It can become a response to skills scarcity.
For global mobility and HR teams, that creates a wider strategic role. The question is no longer only how to relocate a person after someone has been recruited. Organizations may need to determine where particular capabilities can be found, whether those people need to relocate at all, and which employment model makes the talent accessible.
That could mean sponsoring an employee in one market, hiring remotely in another, relocating a high-demand specialist, using an Employer of Record where the organization lacks an entity, or developing a distributed team across several countries.
In other words, the talent shortage can change the geography of the organization itself.
But global hiring cannot create skills that do not exist
There is an important limitation to that strategy.
If employers in the United States, Europe and other major economies all respond to scarce AI capabilities by looking overseas, they can broaden the pool available to any individual company. But they do not necessarily enlarge the global supply of those capabilities.
Companies may simply find themselves competing internationally for the same relatively small group of experienced workers.
The ManpowerGroup findings illustrate why this matters. Talent shortages remain widespread across industries, from information and professional services to manufacturing, hospitality, healthcare and the public sector. Geography changes the intensity of the shortage, but not necessarily the underlying mismatch between the skills employers need and the skills currently available.
That makes talent development more than an HR initiative. It becomes part of the solution to a global supply problem.
Organizations that can teach existing workers to use AI effectively, identify employees capable of moving into emerging roles, and develop inexperienced workers into future specialists may gain an advantage over those that depend predominantly on external recruitment.
The workforce strategy therefore starts to resemble a portfolio: buy some capability from the market, build some internally, borrow some through contingent talent and use technology where appropriate.
And that makes the entry-level question much more important.
The talent companies want tomorrow has to start somewhere
The World Economic Forum’s June 2026 report on AI and entry-level work, developed in collaboration with PwC, estimates that 37% of young workers globally are already employed in occupations with medium-to-high exposure to AI-driven task change.
The exposure is considerably higher in several advanced economies: 69% in Northern America, 63% in Europe and 75% in Eastern Asia.
The issue is not simply whether AI eliminates jobs. The Forum finds a much more complicated transition.
AI can remove routine tasks, allow junior employees to tackle more sophisticated work earlier and increase productivity. Indeed, 68% of entry-level workers surveyed reported productivity improvements from AI.
But those gains can create another problem if organizations automate the tasks through which inexperienced workers traditionally learned.
Junior employees historically developed professional capability by performing basic work repeatedly, observing more experienced colleagues and gradually assuming more complex responsibilities. Some of those tasks may appear inefficient when viewed purely through the lens of automation. Yet they may also have served as training.
Remove too many of them without replacing the learning mechanism, and an organization may discover several years later that fewer employees have developed the judgment, domain expertise and contextual understanding needed for more senior responsibilities.
The report captures the dilemma in one employer observation: organizations can “build, buy, borrow” the talent they require—but if everyone stops building entry-level talent, there will eventually be less talent available to buy.
Entry-level jobs may need redesigning, not protecting unchanged
That does not mean companies need to preserve entry-level jobs exactly as they existed before AI. The more important question may be what those positions are designed to accomplish.
The World Economic Forum finds that organizations achieving stronger AI-related outcomes are more likely to redesign workflows rather than simply place AI tools on top of existing processes. Yet only 16% of organizations cited in the report had fully redesigned roles, processes and operating models around AI.
That leaves considerable room for experimentation.
An entry-level employee who previously spent hours compiling information might instead use AI to accelerate that work and spend more time interpreting results, questioning outputs, communicating findings or participating in decision-making.
But employers must decide deliberately which activities should disappear and which should remain because they develop capability.
The report cites Hitachi, for example, as retaining experiences that build judgment and critical thinking even as repetitive coordination work becomes automated. Early-career engineers are still expected to understand coding fundamentals so they can validate and challenge AI-generated code rather than simply accept its output.
That distinction may become central to workforce planning: automation should remove unnecessary work without accidentally removing the experiences through which expertise is formed.
Skills may matter more than traditional career ladders
AI is also beginning to challenge the assumption that careers must progress through a predictable hierarchy.
The Forum reports that organizations are exploring more capability-based models in which people move across projects and assignments according to what they can do rather than simply how long they have occupied a particular job.
That can create opportunities for employees to reach complex work earlier.
It can also make development less linear.
Instead of progressing from junior employee to manager through a fixed sequence of positions, workers may develop through combinations of assignments, cross-functional exposure, AI-enabled work and continuous learning.
For global mobility, that development could increasingly include international experience without requiring every employee to undertake a traditional long-term assignment. Project deployments, cross-border teams, virtual collaboration, temporary relocation and international rotations can all become mechanisms for developing capabilities.
Mobility therefore has the potential to contribute not only to where companies find talent, but also to how they develop it.
Employers and educators may need a shorter feedback loop
There is another reason simply searching globally will not be enough: skills requirements are changing faster than many education systems can adjust.
The World Economic Forum reports that entry-level occupations with the highest AI exposure are experiencing substantially faster skills change than less exposed positions. Around 28% of entry-level workers believe that half or fewer of their current skills will remain relevant in three years.
Traditional degrees remain important, but employers increasingly value additional evidence of readiness: applied experience, problem-solving, workplace technology exposure and the ability to work effectively with AI.
That strengthens the case for closer connections among employers, universities, training organizations and workforce planners.
Instead of education ending when employment begins, talent development may increasingly become continuous—starting before someone joins an organization and continuing throughout a career.
Companies that know what capabilities they will require can also help shape those pipelines earlier through internships, apprenticeships, university partnerships, certifications and work-integrated learning.
Global talent strategy becomes a build-and-find strategy
Taken together, the three studies suggest that talent strategy is moving in two directions at once.
Organizations will need to find talent more broadly—across countries, employment models and previously overlooked talent pools.
But they will also need to build talent more deliberately—through upskilling, redesigned entry-level positions, internal mobility, international exposure and closer connections with education and training.
That may be particularly important for scarce capabilities such as AI, where today’s shortage cannot be solved indefinitely by companies competing to hire the same experienced people from one another.
For global mobility, this changes the conversation as well.
Mobility can help organizations access talent wherever it exists. But the more strategic opportunity is to become part of the decision about how capabilities are sourced, developed and deployed across markets.
The conversation may therefore evolve from:
Where can we relocate this employee?
to:
Where can we find this capability? Can we develop it internally? Does the person need to move? Could the work move instead? And how do we build a sustainable pipeline so we are not searching for the same scarce skill again three years from now?
Those are workforce strategy questions, not simply relocation questions.
And as technological change accelerates, the organizations with the strongest talent pipelines may not necessarily be those that search the widest.
They may be the ones that understand a more fundamental reality: the global competition for talent cannot be won only by finding people. At some point, somebody has to develop them.