The AI divide is the growing gap between economies where artificial intelligence reshapes work fastest and economies that lack the people and skills to use it at all. IMF staff estimated in January 2024 that about 60% of jobs in advanced economies are exposed to AI, against 40% in emerging markets and 26% in low-income countries. The result is two opposite imbalances: expertise under pressure in one part of the world, and businesses that cannot grow for lack of it in another. BrainGain.Expert was built to connect the two.
If AI takes the jobs, why invest in people?
This week, AI and jobs opened the IMF and World Bank Annual Meetings in Bangkok, with Ajay Banga, Kristalina Georgieva and Singapore President Tharman Shanmugaratnam on stage at the session Small AI: More and Better Jobs on 12 October 2026, according to The Nation Thailand.
Clips from that week traveled fast under headlines like "AI will eliminate 40% of jobs." Soon after, we heard a question that now comes up in almost every conversation about our work: if AI is about to take a large share of work, why fund an initiative that puts experienced people inside businesses?
It is a fair question. Our answer: AI does not weaken the case for expertise inside companies. It moves the problem, and the 40% figure shows where.
Forty percent is exposure, not a layoff forecast
IMF staff found that almost 40% of global employment is exposed to AI, as Kristalina Georgieva wrote on the IMF blog in January 2024. Exposure means that AI can perform or support the key tasks of a job. In advanced economies, roughly half of the exposed jobs may benefit through higher productivity. For the other half, AI may take over key tasks, which can mean less hiring and lower wages, and in the most extreme cases some jobs may disappear.
The number that went viral measures how much work will change, not how many people will lose it.
The headline average hides the most important detail: AI does not arrive everywhere at the same speed. Advanced economies are more exposed because more of their jobs are cognitive: analysts, managers, finance and operations specialists. Poorer economies are less exposed, but the IMF warns that many lack the infrastructure and skilled workforce to benefit, and that AI could widen the gap between countries over time.
Two economies, two opposite problems
In advanced economies, the pressure falls on experienced professionals. In a Staff Discussion Note published in January 2026, IMF economists found that the spread of AI skills is linked to lower employment in occupations that are highly exposed to AI and gain little from it. Some of the people closest to corporate hiring expect the shift to come quickly. Workday CEO Aneel Bhusri, whose software runs HR for many of the largest US companies, believes AI-driven job displacement is coming within five years, Fortune reported on 13 October 2026. Earlier this year he told diginomica that the industry must find a way to take care of the people affected: "We have to come up with a plan for them." Even before AI, expertise was not fully used: in 2024, 21% of employed university graduates in the EU worked in jobs below their qualification level, according to Eurostat. AI is likely to add to that pool, not shrink it.
In fragile and emerging economies, the problem runs the other way. They do not lack people in general. Over the next ten years, 1.2 billion young people across emerging economies will reach working age, while only about 420 million jobs are currently expected to be created, according to the World Bank. What they lack is the experienced people who build the companies that create those jobs: the operations lead, the export manager, the engineer who can bring in new technology. In Ukraine, war has made this shortage acute. In June 2026, a record 71% of industrial firms named staff shortages caused by mobilization or emigration as their main obstacle, according to a monthly survey by the Institute for Economic Research and Policy Consulting.
One part of the world is about to have more expertise than its jobs can absorb. Another cannot grow because it has too little.
Left alone, these two imbalances feed each other: roles shrink on one side, businesses stall on the other. The usual way to close such a gap is migration, which drains the country that can least afford to lose people. Ukraine already has some 5.76 million refugees abroad, according to UNHCR.
Why the technology alone will not close the gap
One might expect AI itself to compensate: if a company lacks people, let software do the work. In practice, the companies that need AI most are the least able to adopt it.
Six months after buying two AI subscriptions, the owner of a 60-employee logistics company in western Ukraine has used neither. She had read that AI could process waybills, customs declarations and supplier invoices in minutes. No one had time to map how documents actually move through the company, connect the tools to the accounting system or train the team. The one employee who might have done it was mobilized.
The pattern holds well beyond Ukraine. In 2025, 55% of large enterprises in the EU used at least one AI technology, against 17% of small enterprises, according to Eurostat. Price is not the main barrier. Among Swedish companies that considered AI in 2025 and decided against it, 74.7% named a lack of relevant expertise as the reason, Statistics Sweden reported. If expertise holds back firms in one of the most digital economies in Europe, a company that has lost its specialists to war starts much further behind.
AI does not remove the need for expertise inside a company. It raises the cost of not having it.
“AI can boost growth & create opportunities, but the gains are not automatic.”
Kristalina Georgieva, Managing Director, International Monetary Fund
Part of the plan: a new address for experience
Bhusri is right that the people whose work AI reshapes need a plan. Most plans so far look inward: retrain people for the jobs that remain at home. That is necessary, but it misses something. The expertise that AI puts under pressure in one economy is often exactly what a company in another economy cannot find.
A finance lead in Munich whose reporting work is being automated has already done what a Ukrainian manufacturer needs: rebuilt a finance function around new tools. An operations specialist in Milan knows how to redesign a warehouse workflow before a single robot arrives. Their careers are being reshaped, but their experience has not lost its value. It has lost its address.
“Everyone agrees we need a plan for the people whose work AI is reshaping. Part of that plan is simple: put their experience to work where it is missing. If expertise goes unused in one place while businesses stall for lack of it in another, both sides lose. BrainGain.Expert connects the two: brain gain without brain drain.”
Oksana Pashchenko, Co-Founder and Co-CEO, BrainGain.Expert
Remote work makes the connection possible without anyone having to move. The professional stays where they live. The company gets the experience it lacks, inside its own team, for as long as it takes to make the change stick. The country keeps its people, and gains capability instead of losing it. The work is real and paid: donors fund the Fellow's project work, and when a Fellow takes a role inside the company, the business pays a market salary, with support from the program in the first months.
When AI reshapes a career in one economy, that experience can rebuild a company in another.
We are not the whole plan. Retraining, social protection and new jobs at home remain essential, and no single platform can rebalance global labor markets. What we can add is a working channel between the two sides, proof that it changes results inside companies, and a model that employers restructuring around AI and public funds for workers in transition can make part of their own plans. The AI divide will not close by itself, but it can be bridged one company and one expert at a time.
Domain expertise first, AI where it pays off
BrainGain.Expert is not an AI integrator, and our model stays the same. Every engagement follows the stages of a BrainGain.Expert program: a diagnosis at entry, a Fellow the business chooses, then Advise and Solve, a project of 4 to 12 weeks to analyze the gap and build a first working version, and where needed Implement and Operate, a role of 12 to 24 weeks inside the company to put the change to work. When the Fellow steps back, the capability stays with the team. Fellows are chosen for what they have done: run a finance function, rebuilt a supply chain, brought a product to a new market. AI is one of the tools they use. The platform makes sure they and the business have everything else they need, in every economy where we work.
- 1. Diagnosis. Every diagnosis maps where the business loses capacity, in people, processes, structure or technology, including which tasks AI could take over or support.
- 2. Matching. We shortlist Fellows with proven expertise in the domain the business needs, from operations and finance to export and engineering, and the business chooses its own.
- 3. Process redesign. In the Advise and Solve stage, the Fellow redesigns the process with the team. In the Implement and Operate stage, they put it to work inside the company, adding AI and other tools only where they pay off.
- 4. Training. The platform prepares both sides for the economy of tomorrow: Fellows learn the local context and the newest tools, and business teams learn to run the new ways of working on their own.
- 5. Resources and partners. Technology partners contribute software, cloud and AI credits to the Resource Pool, and partner organizations add specialist support where a redesign needs more than one Fellow. No partner buys access to the businesses we serve.
Many problems in a recovering company are about people, structure or cash, and a good Fellow knows when a clear owner beats a new platform. But no business should leave our program less ready for the economy it will compete in.
How we will measure it honestly
A role that AI redesigns is not a failed placement, and an installed tool is not a capability. We are adjusting how we count before the pilot starts, not after.
- Balanced redesign: every process redesign curated by the platform must pass two tests at once: the process runs more efficiently, and people keep a meaningful role in it. The goal is better work for the team, not a process with no one left in it.
- People: a role counts as kept when the person is still in qualified work at 6 and 12 months, including when AI has redesigned the role. A role removed by automation is reported, not hidden.
- Capabilities: a tool counts only if it is in daily use three months after handover and has an internal owner.
- Productivity: we add output per employee to the indicators we track for each business, because that is where AI gains show first.
- Additionality: a role or a tool counts only if it would not have been in place without the program.
Results are evaluated independently and published once a year, including what did not work and why.
The gains are not automatic: people make them
AI will reshape work in every economy, but not at the same speed and not with the same consequences. Whether it deepens the divide or helps close it will be decided less in policy speeches than inside thousands of small companies: by whether they can reach experienced people who redesign the work around new tools and keep their teams at the center of it. That is the gap we described in People need people: why we built BrainGain.Expert, and AI makes it both wider and easier to bridge. If you run a business, have expertise to give or can help fund the first cohort, this is the moment to join us.