> **来源：[研报客](https://pc.yanbaoke.cn)** # Summary of Global AI Adoption in 2025 ## Core Content In 2025, global adoption of artificial intelligence (AI) continued to rise, with generative AI tools now used by roughly one in six people worldwide. The growth rate in AI adoption was 1.2 percentage points, reflecting the rapid integration of AI into daily life. However, this progress has been uneven, with a widening digital divide between the Global North and the Global South. ## Main Points - **Global Adoption Growth**: - AI diffusion increased from 15.1% in H1 2025 to 16.3% in H2 2025. - The Global North saw a faster adoption rate (24.7%) compared to the Global South (14.1%), with the gap widening by 0.9 percentage points. - **Regional Leaders**: - The United Arab Emirates (UAE) and Singapore remained the top two countries in AI adoption, with the UAE leading at 64.0% and Singapore at 60.9%. - South Korea made the most significant leap in the rankings, rising seven positions from 25th to 18th due to national policies, improved language models, and a viral cultural moment. - **Factors Driving AI Adoption**: - **National Policies**: South Korea's strategic investments in AI education and infrastructure accelerated its adoption. - **Language Model Improvements**: Enhanced Korean-language performance of models like GPT-4o and GPT-5 made AI more accessible and practical for local users. - **Cultural Impact**: A viral trend of Ghibli-style image generation using ChatGPT-4o sparked widespread interest and adoption in South Korea. - **DeepSeek's Emergence**: - DeepSeek, an open-source AI platform, gained significant traction in regions where Western platforms had limited access, including China, Russia, Iran, and Africa. - Its open-source MIT license and free-to-use chatbot made it accessible to a broader audience, especially in price-sensitive and underrepresented markets. - **UAE's Strategic Leadership**: - The UAE's early and deliberate AI strategy, including the appointment of the world's first Minister of State for AI and the creation of sandbox environments, contributed to its high AI adoption rate and public trust. - **Challenges and Future Outlook**: - The global AI diffusion trend shows a concentration of growth in already digitally advanced economies, raising concerns about the digital divide. - The next wave of AI users may come from the Global South, enabled by open-source innovation and reduced access barriers. ## Key Information - **Methodology**: AI diffusion is measured as the percentage of the working-age population using generative AI tools, based on Microsoft telemetry and adjusted for market share and internet penetration. - **AI Diffusion by Region**: - **Global North**: 24.7% in H2 2025 (up from 22.9% in H1). - **Global South**: 14.1% in H2 2025 (up from 13.1% in H1). - **Top AI Adopting Countries in H2 2025**: - 1. UAE (64.0%) - 2. Singapore (60.9%) - 3. Norway (46.4%) - 4. Ireland (44.6%) - 5. France (44.0%) - 6. Spain (41.8%) - 7. New Zealand (40.5%) - 8. Netherlands (38.9%) - 9. United Kingdom (38.9%) - 10. Qatar (38.3%) - **AI Adoption in South Korea**: - Surged from ~26% to 30.7%, becoming one of the fastest-growing markets for ChatGPT. - The growth was fueled by government policy, improved language models, and a cultural phenomenon. - **DeepSeek's Impact**: - Gained popularity in China, Russia, Iran, Cuba, Belarus, and Africa due to its open-source nature and affordability. - Demonstrated how open-source AI can serve as a tool for expanding access in underserved regions. - **U.S. Adoption**: - Maintained high absolute usage (28.3%) but fell from 23rd to 24th place in global rankings, behind smaller, more AI-focused economies. ## Conclusion The year 2025 saw rapid global AI adoption, with generative AI tools becoming more mainstream. However, the growth was uneven, highlighting a growing digital divide. Countries with early investments in digital infrastructure and AI policies, such as the UAE, Singapore, and South Korea, led the way, while open-source platforms like DeepSeek are reshaping the AI landscape in previously underserved regions. The challenge moving forward is to ensure that AI innovation benefits all regions equally, reducing disparities and expanding access to underrepresented communities.