AI Diffusion Where AI is most used, developed and built (www.microsoft.com)

A new AI Diffusion report argues that AI is the newest general‑purpose technology—and the fastest to spread—reaching more than 1.2 billion users in under three years. But adoption is highly uneven: AI use in the Global North is roughly double that of the Global South, with the gap widening sharply below a GDP per capita of $20,000. Working‑age adoption exceeds 50% in places like the UAE (59.4%) and Singapore (58.6%), while parts of Sub‑Saharan Africa and Asia show rates below 10%. Nearly four billion people still lack the basic electricity, connectivity, and digital literacy needed to use AI, and language barriers further depress uptake in low‑resource language regions. Technically, the report frames diffusion as the product of three forces—frontier builders (model creators), infrastructure builders (compute, data centers, networks) and users—and proposes three indices (AI Frontier, AI Infrastructure, AI Diffusion) to track them. It highlights concentrated infrastructure: the U.S. and China host about 86% of global data‑center capacity, and only seven countries (U.S., China, France, S. Korea, U.K., Canada, Israel) account for top‑200 models, with the U.S. leading but performance gaps narrowing (China ~6 months behind; span from U.S. to Israel ~11 months). The takeaway: without coordinated investment in electricity, data infrastructure, skills and language localization—plus supportive policy—AI’s benefits will remain concentrated, making diffusion (not just invention) the decisive factor for who gains from AI.