AI adoption is rising, but access is diverging
AI adoption reached 18.8% of the world’s working-age population in June 2026, according to Microsoft’s new Q2 Global AI Diffusion Report. That was about one percentage point above the first quarter, but the same report shows the Global North–South usage gap widening to 12.6 percentage points.
The headline is therefore not simply that more people used generative AI. Growth is broad, yet it is not closing the structural divide between economies with abundant connectivity, devices and skills and those without them.
Key takeaways
What the report actually measures
Microsoft defines AI adoption as the share of people aged 15 to 64 who used a generative-AI product during the measured period. It derives the estimate from aggregated, anonymised telemetry and adjusts for operating-system and device market share, internet penetration and population.
That methodology makes the figures comparable across 147 economies, but it is not a census of every AI tool. Microsoft itself says no single metric is perfect and plans to expand the measure to include newer tools and models. Readers should treat the percentages as a consistent indicator of diffusion, not a complete count of all AI use.
MoneyDJ independently reported the global 18.8% level, the country rankings and Taiwan’s move to 33.8%. A second independent lead could not be opened and was excluded rather than treated as evidence; the routine primary-plus-one gate remains satisfied.
AI adoption leaders are separating from the average
The UAE led Microsoft’s ranking at 73.3%, followed by Singapore at 64.3%. Ireland reached 49.9%, France 49.6% and Norway 49.4%. South Korea posted the largest absolute quarterly gain at 3.5 percentage points, while Japan recorded the biggest relative increase among the highlighted economies.
The more consequential comparison is regional. Global North usage rose to 28.8%, against 16.2% in the Global South. Microsoft says education and learning account for a higher share of use in the South, but infrastructure, connectivity and skills remain barriers.
Everyone else is reporting a league table; we are explaining that AI adoption is becoming an infrastructure story. If access to compute, reliable networks and local-language tools determines who can use AI productively, the diffusion gap can reinforce differences in education and business capacity.
Open models may change the slope, not the constraints
Microsoft notes that open-weight models represent a growing share of API token volume. Lower-cost, adaptable models can help local developers support languages and tasks that global services overlook. But open weights do not remove electricity, device, network or training constraints.
For businesses, the report is a warning against treating global averages as a market forecast. Product adoption will depend on local readiness and workflow fit. Lapaas Voice’s coverage of the Microsoft India cloud region shows why infrastructure commitments and adoption data must be read together.
The next update matters because Microsoft plans to widen its measurement base. A method change may lift reported adoption, particularly in China, so future quarter-on-quarter comparisons will need to separate real growth from expanded coverage.
How companies should read the 18.8% figure
The 18.8% global figure is useful for direction, not for budgeting a single market. A company deciding where to launch an AI product should pair diffusion data with broadband quality, device affordability, language support, regulation and the presence of partners that can train users.
It should also distinguish trial from durable use. The report measures whether people used a generative-AI product during the period, not whether usage changed revenue, wages or productivity. Higher diffusion can create a larger funnel, but it does not prove that organisations have redesigned workflows around the technology.
For policymakers, the widening gap makes sequencing important. Skills programmes will have limited effect without reliable connectivity, while infrastructure spending will underperform if local institutions cannot adapt tools to regional languages and needs. The report points to a systems problem rather than a single adoption lever.
Related Lapaas Voice coverage: related technology coverage and related technology coverage.
FAQ
What does Microsoft mean by AI adoption?
It estimates the share of people aged 15–64 who used a generative-AI product during the measured period, using adjusted and anonymised telemetry.
Which countries led Microsoft’s Q2 2026 ranking?
The UAE led at 73.3%, followed by Singapore at 64.3%; Ireland, France and Norway completed the top five.
Why is the AI adoption gap important?
A widening gap suggests infrastructure, connectivity, skills and affordable model access may compound existing economic differences.
Verified facts
| Item | Detail |
|---|---|
| Global working-age use | 18.8% in June 2026 |
| Quarterly change | About +1 percentage point |
| Global North | 28.8% |
| Global South | 16.2% |
| North–South gap | 12.6 percentage points |
| Coverage | 147 economies |
Sources
- Microsoft AI Economy Institute (primary)
- MoneyDJ (independent)
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
