Global AI Usage Climbs Beyond Early Experimentation
Global AI usage is rising steadily, signaling a shift from experimentation to mainstream adoption. According to Microsoft’s latest Global AI Diffusion report, the share of the world’s working-age population using generative AI rose from 16.3% to 17.8% in the first quarter. Twenty‑six economies now report more than 30% of their working-age citizens using AI tools, reflecting rapid diffusion in certain markets. At the top of Microsoft’s National AI Leaderboard, one Gulf nation leads with 70.1% usage, while other high performers include city-states and European economies above 40%. Even traditionally slower movers are climbing the ranks as AI becomes embedded in everyday tasks like messaging, search, learning, and content creation. At the same time, AI coding platforms are accelerating software development, with global Git pushes up 78% year over year and new repositories up 45%, underscoring how deeply AI is seeping into technical workflows and productivity.
The Widening AI Adoption Gap Between North and South
Beneath the headline growth figures lies a stark reality: the AI adoption gap between the Global North and South is widening. Microsoft’s data shows that 27.5% of the population in the Global North used generative AI in the latest quarter, compared with just 15.4% in the Global South. While both groups are growing, adoption in the North is expanding more than twice as fast, pushing the gap from 10.6 to 12.1 percentage points. This divergence points to a structural digital divide in AI, not simply a lag that time will fix. The report links slower uptake in many emerging economies to basic infrastructure constraints—unreliable electricity, patchy internet connectivity, and lower digital literacy. These factors turn AI adoption into a broader education and workforce challenge, rather than a simple matter of making tools available or lowering prices.
Language, Skills, and Infrastructure: What Drives Uneven Adoption
The uneven geography of global AI usage is shaped by more than income levels. Infrastructure, language support, and workforce capabilities are emerging as decisive factors. Regions with reliable broadband and stable electricity grids are better positioned to integrate AI into everyday workflows. Meanwhile, an AI skills shortage in many developing economies limits how effectively organizations can adopt and adapt these tools. Microsoft highlights the role of language in accelerating diffusion: as generative AI systems improve in Asian languages and support multimodal interaction, adoption has surged in multiple Asian markets. Several economies in the region rank among the 15 fastest‑growing AI adopters, with some recording user-share increases above 30% since mid‑2025. This suggests that when tools reflect local languages and contexts—and when basic infrastructure is in place—AI usage can climb quickly, even outside traditional technology hubs.
Enterprise Demand and Coding Tools Concentrate Benefits
Enterprise adoption is amplifying the AI adoption gap. Large organizations in advanced economies are pouring AI into software development, knowledge work, and productivity tools. AI coding assistants from major providers have contributed to a 78% year‑over‑year rise in Git pushes and a 45% increase in new Git repositories, signaling an explosion of AI‑augmented software projects. Merged pull requests linked to AI coding agents have grown more than 28‑fold since mid‑2025, illustrating how quickly professional developers are building AI into their workflows. Early labor data suggests that, at least for now, AI is complementing rather than replacing software developers, with employment in that profession continuing to grow. By contrast, firms in many emerging markets lack the infrastructure, capital, and skilled talent to replicate these gains, reinforcing a two‑speed AI economy where productivity boosts accrue disproportionately to already advanced sectors and regions.
Economic Risks of a New Digital Divide in AI
If current trends continue, the digital divide in AI risks becoming a structural economic fault line. Countries and firms that harness AI to boost productivity, expand software development, and upgrade services could see compounding advantages in innovation, competitiveness, and wages. Meanwhile, regions constrained by weak infrastructure and limited digital skills may struggle to capture similar gains, deepening existing income and opportunity gaps. The AI adoption gap could translate into an AI productivity gap, where entire industries in lower‑adoption economies fall further behind. This is not only a question of tool access, but of whether education systems, training programs, and connectivity policies evolve fast enough to support broad‑based AI literacy. Without deliberate investment in skills and infrastructure, billions risk being positioned primarily as AI consumers rather than AI creators, entrenching a new hierarchy in the global digital economy.