• Microsoft data shows AI use rose in late 2025, save in the US.
• Adoption remains uneven due to trust and access.

AI use kept rising through the second half of 2025, but the gains were uneven. In the United States and much of Western Europe, more people are using generative AI tools at work, in school, and at home. At the same time, the data shows that adoption is spreading faster in smaller, highly digitised countries than in the US, even though American firms still shape much of the technology itself.

By the end of 2025, about one in six people worldwide had used a generative AI product. That figure rose by just over one percentage point in six months, a solid increase for tools that only entered daily life a few years ago. The findings come from Microsoft’s AI Diffusion Report 2025, a study by Microsoft that tracks how many people used a generative AI product during the period, based on aggregated and anonymised data adjusted for population size, internet access, and device mix.

In the report, Microsoft links uneven AI adoption to differences in trust, access, and how easily people can use AI tools in daily life.

What stands out most is not the global average, but where growth is happening. Across North America and Western Europe, AI use continues to climb, yet it is not accelerating at the same pace as in some smaller economies that have focused early on public-sector use, digital skills, and everyday access.

The United States builds the tools, but adoption lags

In the United States, AI is everywhere in headlines, products, and investment plans. The country still dominates model development and computing infrastructure. Yet when measured as a share of the working-age population, US use sits below several European peers. By the end of 2025, about 28% of working-age Americans were using generative AI tools. That placed the US in the mid-20s globally, down one spot from earlier in the year.

The gap points to a basic tension. Strong research labs, venture funding, and cloud capacity do not guarantee widespread adoption in society. Many Americans still see AI as something used by specialists, students, or office workers in certain fields, rather than as a general-purpose tool. Concerns about data use, accuracy, and job impact remain higher in the US than in other European countries.

The result is wide exposure but uneven depth. People may try AI once or twice, but few use it regularly for everyday tasks. Although innovation continues at a rapid pace, it does not always translate into daily habits.

Europe shows steadier, more consistent uptake

Western Europe presents a more diverse but frequently steadier picture. Countries like France and Spain continue to rank near the top in overall use, thanks to public-sector trials, education programmes, and clearer national guidance. Others, including the United Kingdom and parts of Northern Europe, show gradual but slower gains.

The top of the rankings has become crowded, and most changes in the top 30 are small. That stability demonstrates how saturated adoption has become in countries that invested early. In many cases, AI is already integrated into government services, classrooms, or workplace tools, leaving little room for sharp jumps.

South Korea, on the other hand, rose rapidly in the world rankings by the end of 2025. While not a Western economy, its experience demonstrates what faster adoption can look like. Policy coordination, better language support in models, and consumer-facing features made AI feel practical in daily life. Use rose above 30% of the population in a short period.

By comparison, Western countries, especially the US, have taken a more fragmented path. Agencies, schools, and companies all move at different speeds, frequently without shared standards. That flexibility supports experimentation, but it can slow broad adoption.

Trust, access, and what comes next

Trust remains a key factor. According to surveys conducted in 2025, Americans remain cautious about AI, particularly when it comes to accuracy and misuse. That caution does not prevent experimentation, but it does influence how deeply people rely on these tools. In countries where governments introduced AI earlier in public services, adoption frequently feels more routine and less risky.

Another shift affecting the US and Europe is competition from open and low-cost systems developed elsewhere. While most Western users still rely on domestic platforms, the global rise of free tools shows how pricing and access influence adoption. In wealthier markets, these alternatives have had limited impact so far, but they underline a broader point: use is driven as much by convenience and trust as by technical strength.

Taken together, the second half of 2025 shows steady progress rather than a surge in the United States and Western Europe. More people are trying AI, but fewer are making it central to how they work or learn. The Microsoft data suggests leadership in research and infrastructure must be paired with simpler access, clearer rules, and visible public benefits if AI adoption is to accelerate.

AI adoption is about who helps people trust them, understand them, and find practical reasons to use them. For the US and its Western peers, the next phase may depend less on technical breakthroughs and more on whether AI becomes an ordinary tool, rather than a headline feature.

Want to experience the full spectrum of enterprise technology innovation? Join TechEx in Amsterdam, California, and London. Covering AI, Big Data, Cyber Security, IoT, Digital Transformation, Intelligent Automation, Edge Computing, and Data Centres, TechEx brings together global leaders to share real-world use cases and in-depth insights. Click here for more information.

TechHQ is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.