Global AI diffusion is rising, but not evenly
New data from Microsoft’s Global AI Diffusion Report shows that AI usage is moving beyond early experimentation. In the first quarter of 2026, AI diffusion – defined as the share of people aged 15 to 64 using generative AI – rose 1.5 percentage points, from 16.3% to 17.8% of the world’s working-age population. Twenty-six economies now report more than 30% of their working-age populations using AI tools. Leaders include the UAE at 70.1%, with other high-usage economies such as Singapore, Norway, Ireland, and France also surpassing the 40% mark. Even traditionally slower movers are climbing: the United States, for example, advanced in Microsoft’s national leaderboard as its usage rate reached 31.3%. These figures underline a clear trend: AI is rapidly diffusing into everyday work and life, but the geography of this growth is highly uneven and increasingly polarized.
A widening AI adoption gap between Global North and South
Behind the headline growth numbers lies a more troubling pattern: the AI adoption gap between the Global North and South is widening. In Q1 2026, generative AI usage in the Global North climbed to 27.5% of the working-age population, up from 24.7% in late 2025. The Global South increased from 14.1% to 15.4% over the same period. This pushed the difference between the two blocs from 10.6 to 12.1 percentage points. Microsoft links this divergence to foundational factors such as reliable electricity, robust internet connectivity, and baseline digital skills – not simply access to software. In other words, global AI skills are growing fastest where digital infrastructure and education systems are already strong, reinforcing existing inequalities. Without targeted interventions, today’s AI diffusion rates risk hardening into a long-term structural divide in productivity, innovation capacity, and job opportunities.
Skills, infrastructure, and enterprise AI adoption
The uneven spread of AI is closely tied to differences in training, infrastructure, and enterprise AI adoption. Regions with higher diffusion typically combine widespread broadband access, reliable power, and a digitally literate workforce. This environment allows organisations to integrate AI into workflows, invest in experimentation, and build internal capabilities. Microsoft’s telemetry-based metrics suggest that the share of people using generative AI correlates with corporate resources like cloud platforms, security, and data governance – areas where large enterprises in developed markets are more advanced. In contrast, organisations in less connected regions often struggle with bandwidth constraints, inconsistent access to devices, and limited budgets for upskilling. The result is a dual-speed world: in one track, AI becomes embedded in everyday work; in the other, workers remain stuck at the experimentation fringe, with little opportunity to develop durable global AI skills.
Regional contrasts and the role of language
Regional patterns show that the AI adoption gap is not simply a North–South story. Some economies are rapidly catching up by addressing linguistic and usability barriers. Asia stands out: twelve of the fifteen fastest-growing AI adopters since mid-2025 are in the region, with South Korea, Thailand, and Japan recording the largest increases. Japan’s AI user share climbed 3.4 percentage points in a single quarter, more than three times the global average, helping it move up the national rankings. Microsoft credits improvements in non-English language performance and multimodal capabilities, which make AI tools more accessible for messaging, search, learning, and content creation in local languages. Enhanced performance on Japanese professional exams and benchmarks such as MMLU underscores how language-aware models can unlock new professional use cases, narrowing the AI adoption gap for users previously constrained by English-centric systems.
Why targeted AI education and infrastructure matter now
The divergence in AI diffusion has clear implications for policy and business strategy. As generative AI becomes embedded in coding, productivity tools, and knowledge work, the risk is that workers and firms in under-resourced regions fall permanently behind. Microsoft’s data on surging software-development activity, enabled by AI coding tools, illustrates how quickly value can accumulate where skills and infrastructure are already in place. To avoid locking in a structural AI adoption gap, governments and enterprises need coordinated investments in digital infrastructure, foundational digital literacy, and targeted AI training. Small and midsize businesses, especially outside leading economies, will require support to experiment with enterprise AI adoption without prohibitive risk. The lesson from current AI diffusion rates is clear: technology alone will not close the divide; only sustained efforts in education, connectivity, and organisational capacity can ensure a broader share of the world benefits.