2026年第一季度全球人工智能扩散趋势与洞察

Executive Summary Notable developments in the quarter included acceleratingAI adoption in Asia driven in part by improving AI capabilitiesin Asian languages. South Korea, Thailand, and Japan sawthe greatest movement. More broadly, the quarter broughtcontinued widening of the AI gap between the Global North The global adoption of artificial intelligence continued torise in the first quarter of 2026. During the quarter, AI usageincreased by 1.5 percentage points from 16.3% to 17.8%of the world’s working age population. Intensity of useamong economies with the highest rates of AI diffusion also At the top of Microsoft’s National AI Leaderboard, the UAEcontinued to lead global AI diffusion at 70.1%. The UnitedStates finally started to move up the national rankings, albeitonly from 24th to 21st based on a 31.3% usage rate by the Interestingly, the quarter brought added evidence that,at least for now, AI coding capabilities may be increasingdemand for the employment of software developers.As discussed in more detail below, when developerproductivity increases, the cost of building softwaredeclines. If demand for software is elastic, organizationscan respond by building more software across a widerrange of use cases, including across broader economic To track all these trends, we continue to measure AI diffusionas the share of people worldwide between ages of 15and 64 who have used a generative AI product during thereported period. This measure is derived from aggregatedand anonymized Microsoft telemetry and adjusted to No single metric is perfect, and this one is no exception.Through the Microsoft AI Economy Institute, we continueto refine how we measure AI diffusion globally, includinghow adoption varies across countries in ways thatbest advance priorities such as scientific discovery and Sectorally, the quarter saw strengthened AI codingcapabilities leading to a dramatic increase in productionof software code. This was reflected in production byAnthropic’s Claude Code, the OpenAI’s Codex, andMicrosoft’s GitHub Copilot. Git pushes – through which AI adoption in the Global North outpaces the Global South At the same time, the gap between the Global North andGlobal South continues to widen, with adoption in the GlobalNorth growing more than twice as fast as in the Global South.In the first quarter of 2026, 27.5 percent of the population inthe Global North used generative AI, up from 24.7 percent in This widening divide reflects the systemic challenges facingthe Global South, where limited access to reliable electricity,internet connectivity, and digital skills continues to constrainadoption. Until these foundational gaps are addressed, the A New Growth Wave in Asia In addition to rising AI diffusion among the world’s topeconomies, Asia is experiencing a broader growth wave.Twelve of the fifteen fastest-growing economies sinceJune 2025 are in Asia, and each has at least 25% more AIusers than in June 2025. Growth has been led by South similarly strong gains in Mongolia, Iran, Laos, and Turkey(all 30%+). The pattern spans both advanced and emergingeconomies, including Kazakhstan, Kyrgyzstan, Uzbekistan, Fastest growing economies since June 2025 % increase in AI user share, Q1 2026 vs. H1 2025 Local Language and MultimodalCapability Italian, Japanese, Korean, Portuguese, Swahili, Yoruba,and Chinese), have made AI tools increasingly capableof handling multilingual tasks, making them moreaccessible for everyday use cases such as messaging,search, learning, and content creation. Combined with One of the key drivers of the surge appears to be thestronger support for local languages and multimodalinteraction that has expanded the relevance of AI acrossdiverse user groups. Improvements in non-Englishlanguage performance, as measured by multilingual How LLMs perform: English compared to 14 languages Strong Demand and Rapid Scaling than the global average. This acceleration in diffusion isconsistent with improvements in model capabilities and theemergence of a more diversified AI ecosystem, alongside High user demand and a rapid shift from experimentation toreal-world usage are reinforcing this growth. Research fromMcKinsey[10]shows that AI adoption in Southeast Asia isgrowing faster than the global average, with a significantshare of organizations moving beyond pilot use cases to Performance on Japanese professional exams has improvedmarkedly across successive model generations, increasingfrom approximately 50.8% accuracy in earlier modelsto over 90% in recent systems. This represents a stepchange in the ability of AI systems to handle complex,domain-specific tasks in Japanese.[12][13][14][15]Theseimprovements are also reflected in standardized Japanese Japan Accelerates Japan’s position in the global AI diffusion rankingimprovedfrom 56th in H1 2025 to 48th in Q1 2026.Over the past quarter, adoption in Japan increased 3.4percentage points, which is more than three times faster English score.[16][17]