How automation affects workers differently around the world

Existing automation measures treat tasks and occupations the same globally, obscuring how different countries experience technological change. This work creates a country-specific task-based framework spanning 124 countries and 2.33 million task-country labels (99% of global population and GDP). Key findings: automation exposure rises sharply with income but varies widely within income groups; low-income countries face more labor-substituting automation while middle-income countries show mixed effects; simpler automation dominates in poor countries while complex channels and AI scale with wealth; women face disproportionate substitution risk. The atlas disaggregates exposure by automation type, technology channel, and AI involvement—enabling cross-country comparisons and policy analysis tied to development stage.