Nearly 40% of Jobs Face AI Impact

Hands typing on laptop with AI and tech icons overlay — Photo: Deemerwha studio / Shutterstock

The head of the International Monetary Fund warned that artificial intelligence could replace jobs and push inequality higher across and within countries — unless governments act fast.

Story Highlights

The International Monetary Fund says nearly 40% of jobs globally face AI impact, with 60% exposure in rich nations.
Managing Director Kristalina Georgieva warns AI could deepen inequality without strong policy guardrails.
World Bank research counters that AI can lift developing economies if digital divides are bridged.
Both institutions agree policy speed will decide whether AI spreads gains or locks in a new divide.

IMF’s Warning: Big Productivity, Bigger Gaps If Unchecked

International Monetary Fund leaders say artificial intelligence is reshaping the global economy. They also say the gains may cluster in a few countries and firms. The fund estimates that almost 40% of jobs worldwide could be affected by AI, with about 60% of jobs in advanced economies exposed. That mix could boost growth where skills and data centers already exist, and leave others further behind if policy does not adapt in time.

Managing Director Kristalina Georgieva has tied the risk to recent history. She said leaders underestimated the backlash to past waves of globalization. Factories closed, towns lost work, and politics turned bitter. She does not want a repeat with AI. She says AI could “replace jobs and deepen inequality” and in many outcomes “worsen overall inequality” if governments fail to prepare workers and protect those at risk.

Why Exposure Is Highest Where Capacity Is Strongest

International Monetary Fund staff map exposure to the overlap between AI tools and job tasks. Advanced economies employ more people in jobs with cognitive tasks. These workers see more direct AI exposure and also have better chances to benefit if AI complements their skills. Lower income countries face a different bind. They have fewer high-skilled jobs today and weaker digital networks, so they risk slower adoption and fewer early gains.

The fund also flags “frontier” concentration. Training large models needs vast computing power, top talent, and huge datasets. Those inputs sit in a handful of firms and a few countries. That raises the odds of winner-take-most markets and policy capture. It also makes poorer countries dependent on outside platforms and prices. The fund says this dynamic can widen resilience gaps between leaders and laggards.

Counterpoint: AI As a Lifeline If Barriers Fall

The World Bank argues that artificial intelligence can be a lifeline for developing economies if access improves. Its research finds that the share of jobs that could see meaningful productivity gains from AI is similar across economies: about 16.2% in developing countries and 18.7% in high income countries. The bank says the biggest promise is to amplify workers, not replace them, when training and infrastructure are in place.

World Bank materials add that artificial intelligence can cut language and distance barriers. Firms can move more cognitive tasks across borders. That could help poorer countries sell services and climb value chains faster. But the bank also warns that without deliberate action, AI could still widen country gaps, raise inequality within nations, and concentrate market power in a few hands.

What This Means For Workers And Policy Now

For workers in advanced economies, the near-term risk is task change and job churn. International Monetary Fund and related studies say college-educated workers may adapt faster and gain. Older workers and those without training face more strain. For low income countries, the bigger risk is being locked out of early gains due to weak networks, high costs, and low access to cloud computing and skills.

JUST IN – AI ‘widening economic inequality’, warns IMF boss https://t.co/SBNwh5eErF

— Insider Paper (@TheInsiderPaper) October 7, 2026

Both institutions land on the same hinge: policy speed and design. They point to digital infrastructure, open and fair access to tools, skills training at scale, and competition rules that check market power. Clear worker support and safety nets can smooth shocks. Leaders failed to do this in past waves, which fueled distrust. Doing it now could spread benefits more widely and ease the anger that comes when people feel the system is rigged.

How This Fits America’s Broader Debate

In the United States, this warning taps a shared worry on the right and left. Many see elites building wealth while regular people lose ground. Artificial intelligence could speed that split if a few firms control the tools and most gains. Others see a chance to cut costs, raise wages, and bring work to more towns if policy opens access and builds skills. The difference will be choices on training, fair markets, and secure data use backed by results, not slogans.

Limits Of The Evidence And What To Watch Next

Much of the current case rests on exposure estimates, not measured inequality outcomes. The International Monetary Fund is clear that forecasts are not final proof. The World Bank calls the path two-sided as well. Watch for country data tying artificial intelligence adoption to wage changes and firm profits, and for signs that compute access is spreading. Also watch whether developing countries gain service exports as language tools improve.