Global AI adoption reaches 18.8% as North-South divide widens

Generative AI usage among the world's working-age population reached 18.8% in June 2026, marking a one percentage point increase from the first quarter, according to Microsoft's latest Global AI Diffusion Report released today.

The growth proved nearly universal, with adoption climbing across almost every economy tracked. The United Arab Emirates maintained its position at the top with 73.3% usage among its working-age population, followed by Singapore at 64.3%.

Regional leaders and fastest movers

South Korea recorded the largest absolute gain, adding 3.5 percentage points during the quarter. Saudi Arabia made the most dramatic climb in rankings, jumping five positions from 30th to 25th place globally.

Japan demonstrated the strongest relative growth, expanding its user base by 2.2 percentage points—a roughly 10% increase from its Q1 baseline of 22.5% adoption.

Why it matters

The widening gap between developed and developing economies signals a potential bifurcation in how AI's economic benefits will be distributed globally. With usage now at 28.8% in the Global North versus 16.2% in the Global South—a 12.6 percentage point spread—policymakers and technology leaders face mounting pressure to address infrastructure and access barriers before these disparities become entrenched.

Measuring adoption across borders

Microsoft derives its diffusion metric from aggregated, anonymized telemetry data covering people ages 15 to 64 who used generative AI products during the reporting period. The company adjusts raw figures to account for variations in operating system market share, device penetration, internet access, and population demographics.

The methodology, detailed in Microsoft's AI Diffusion technical paper, represents what the company calls "the strongest cross-country measure available today," though officials acknowledge no single metric captures the full picture.

For its next report, Microsoft plans to expand its measurement framework to include newer AI tools and models that have emerged over the past year. The company expects this methodological update to increase reported usage across nearly all economies, with particularly large projected increases in China.

Usage patterns reveal education focus in developing nations

Analysis of Microsoft Copilot usage patterns reveals distinct differences in how AI tools are being applied across regions. The Global South shows a notably higher proportion of education and learning-related use cases compared to the Global North.

Meanwhile, open-weight AI models are capturing growing market share, particularly visible through application programming interface token volume. Microsoft suggests this shift could particularly benefit developing economies by reducing access costs and enabling adaptation to local languages and specific regional needs, though infrastructure, connectivity, and skills gaps remain significant obstacles.

The findings were first reported by Microsoft in its Global AI Diffusion Report published through the company's blog.