What is AI and inequality?

At a glance

• The IMF estimates AI affects roughly 40% of jobs globally, rising to about 60% in advanced economies and 26% in low-income countries.^[2]
• Direction is not fixed: AI widens gaps if it mainly boosts high earners, but some studies show it lifts lower-skilled workers most, shrinking wage gaps.^[4]
• A global divide is forming as capital and capability concentrate in AI-ready countries, leaving less-prepared economies behind.^[5]
• Outcome depends on adoption choices, training, and access more than on the technology itself.^[3]

Within a workforce

AI can compress or stretch pay gaps. When it complements high earners by substituting for clerical tasks, inequality rises.^[1] But research finds that within roles like support, law, and consulting, less-experienced workers often gain the most productivity, narrowing gaps.^[4] For a business owner, who you train decides which way it tilts.

Across countries and firms

Advanced economies and well-capitalized firms are best placed to capture AI gains, while low-income countries lack the infrastructure and skills to adopt fast.^[2] The IMF warns this could widen the global digital divide as capital flows toward AI-ready, regulation-clear jurisdictions.^[5]

Bottom line

AI does not automatically raise or lower inequality; the outcome hinges on who gets the tools, the training, and the gains.