The countries where the most people use AI, ranked

The places that build artificial intelligence and the places that use it most are not the same. The U.S. is home to OpenAI, Anthropic, Google $GOOGL, Microsoft $MSFT, Meta $META and Nvidia $NVDA. Yet it ranks No. 21 in the world when economies are compared by the share of their people who actually use generative AI. The top of that table belongs to smaller, wealthy, densely connected economies — a Gulf state, a Southeast Asian city-state and a cluster of European countries.

Those rankings come from Microsoft’s latest report on the state of global AI diffusion, published Sept. 21 by the company’s AI Economy Institute. The report measures AI adoption by country as the share of people aged 15 to 64 who used a generative AI product during the period. It builds that estimate from anonymized Microsoft telemetry. It then adjusts the estimate for differences in device and operating-system market share, internet access and population.

The headline figure: 18.8% of the world’s working-age population used generative AI in June 2026, up from 17.8% in the first quarter. Growth showed up in almost every economy Microsoft tracks. The gap between rich and developing regions kept widening, though. Usage reached 28.8% in the Global North and 16.2% in the Global South, a spread of 12.6 percentage points.

At the very top, the United Arab Emirates and Singapore are the only economies where more than half of working-age adults use AI. Below them, the race is tight. Ireland, France and Norway are separated by half a percentage point. South Korea posted the largest gain of any economy in the quarter, and Saudi Arabia made the biggest jump in rank.

Microsoft calls its metric imperfect, and it is worth reading with that in mind. It counts how many people use AI, not how often or how intensively they use it. It also does not yet include several newer tools. The company says it will widen the set of tools it measures in its next report. It expects that change to lift measured usage almost everywhere, especially in China.

Here are the 25 economies with the highest share of AI users, in Microsoft’s order.

David Rodrigo / Unsplash

The United Arab Emirates ranks No. 1 in Microsoft $MSFT’s Global AI Diffusion report, with 73.3% of its working-age population using generative AI in the second quarter of 2026. The next-closest economy, Singapore, trails it by nine percentage points.

That figure is up from 70.1% in the first quarter, a gain of 3.2 points. Only South Korea added more over the same three months. For an economy that was already in first place, that pace widened the UAE’s lead rather than simply preserving it.

The UAE’s share is close to four times the global rate of 18.8%. Nearly three in every four people aged 15 to 64 in the federation used a generative AI product during the period Microsoft measured.

The result follows years of state policy focused on AI. The UAE named a minister of state for artificial intelligence in 2017, one of the first cabinet posts of its kind anywhere. It released a national AI strategy with a 2031 horizon the same year. In 2019, Abu Dhabi opened the Mohamed bin Zayed University of Artificial Intelligence, a graduate research university devoted entirely to the field.

The country’s population mix plays a role too. Foreign nationals make up the large majority of residents, and many of them work in finance, aviation, logistics and professional services. Those are sectors where office software and digital tools are part of daily work. The UAE also has some of the highest internet and smartphone penetration rates in the world. That removes a basic barrier that holds down adoption elsewhere.

One more piece of context matters. Microsoft’s figures come from its own product telemetry, and the company has a commercial stake in the region. In 2024, Microsoft invested $1.5 billion in G42, an Abu Dhabi AI company. The report applies the same adjusted, telemetry-based method to every economy it ranks.

The UAE topped the first-quarter table as well. With the June figures, it remains the only economy above 70%.

Hu Chen / Unsplash

Singapore ranks No. 2 in Microsoft $MSFT’s Global AI Diffusion report, with 64.3% of its working-age population using generative AI in June 2026. It and the UAE are the only two economies in the world where more than half of working-age adults use AI.

The city-state held the same rank in the first quarter, when its share stood at 63.4%. Its quarterly gain of 0.9 percentage points was modest next to the UAE’s, but it kept a wide cushion over the rest of the field. Ireland, in third place, trails Singapore by 14.4 points.

Singapore’s AI usage rate is about 3.4 times the global average of 18.8%.

Government planning has shaped that figure over more than a decade. Singapore launched its Smart Nation initiative in 2014 to digitize public services. It set up AI Singapore, a national research and talent program, in 2017, and published a national AI strategy in 2019. In late 2023, it released a second version of that strategy focused on computing capacity, talent and adoption across industry.

The economy itself favors heavy use. Singapore leans on finance, shipping, trade and professional services. Many global banks and technology companies run their Asia-Pacific headquarters from the island. A large share of the workforce does office-based work with corporate software and cloud tools.

Size helps as well. With roughly six million residents on a single main island, Singapore has dense broadband and mobile coverage and a highly educated workforce. English is the main language of business and government. Leading AI chatbots have generally performed best in English, which removes a barrier many other economies still face.

The No. 2 rank also makes Singapore the top economy in East and Southeast Asia by this measure, by a wide margin. South Korea, the next-highest economy in that region, sits at No. 12 with a share 23.7 points lower. Japan, which recorded the fastest relative growth of any economy in the quarter, stands at 24.7%. That puts Singapore’s adoption rate at more than two and a half times Japan’s.

Sophie Popplewell / Unsplash

Ireland ranks No. 3 in Microsoft $MSFT’s Global AI Diffusion report, with 49.9% of its working-age population using generative AI in June 2026. It climbed from No. 4 in the first quarter and is now the highest-ranked economy in Europe.

Ireland’s share rose from 48.4% to 49.9%, a gain of 1.5 percentage points. That was enough to pass Norway, which held third place in the first quarter. The margins at this level are thin. France sits just 0.3 points behind Ireland, and Norway 0.5 points behind.

Ireland’s AI usage rate is more than two and a half times the global rate of 18.8%. Roughly one in every two people aged 15 to 64 in the country used a generative AI product during the quarter.

The country’s economic structure helps explain the figure. Dublin hosts the European headquarters of Google $GOOGL, Meta $META, Microsoft and LinkedIn, among many other U.S. technology companies. Apple $AAPL’s European base is in Cork. Those employers have drawn a large, young, highly educated workforce into software, sales, operations and support roles where digital tools are central.

Ireland’s corporate tax regime has been a major factor in attracting those firms. The country’s 12.5% headline rate drew multinationals for decades. Since 2024, Ireland has applied a 15% minimum rate to large multinational groups under the OECD’s global tax agreement. The cluster of tech employers it built remains.

That cluster has a physical footprint as well. Ireland has become one of Europe’s main data center locations. Data centers used around a fifth of the country’s metered electricity in recent years, according to Ireland’s Central Statistics Office.

Ireland also has one of the youngest populations in the European Union and is an English-speaking country. Both traits tend to go along with faster uptake of new consumer technology. With a population of about 5.4 million, it is a small economy that ranks well above much larger European neighbors on AI use, including Germany at No. 23.

Anthony Delanoix / Unsplash

France ranks No. 4 in Microsoft $MSFT’s Global AI Diffusion report, with 49.6% of its working-age population using generative AI in June 2026. It rose from No. 5 in the first quarter, passing Norway.

France’s share climbed from 47.8% to 49.6%, a gain of 1.8 percentage points. That was the largest quarterly increase among the six European countries in the global top 10. It now trails Ireland by just 0.3 points.

The ranking stands out because of France’s size. With roughly 68 million people, it is by far the most populous economy in the top five. The UAE, Singapore, Ireland and Norway each have populations of about 11 million or fewer. Reaching near-50% adoption across a large and varied national population is a different task from doing so in a city-state.

France has also become one of Europe’s main centers for AI development. Researchers who previously worked at Google $GOOGL DeepMind and Meta $META founded Mistral AI in Paris in 2023. It is one of the best-funded AI companies outside the U.S. and China. It has released several open-weight models, the category of AI system that Microsoft’s report highlights as a growing share of developer usage.

Paris has deeper roots in the field as well. Meta opened a Paris branch of its fundamental AI research lab in 2015. Station F, a large startup campus in the city’s 13th arrondissement, houses many young AI companies. French universities and grandes écoles produce a steady supply of mathematicians and engineers who have gone on to work at major AI labs.

Public policy has backed that ecosystem. In February 2025, Paris hosted the AI Action Summit, a gathering of world leaders and technology executives. President Emmanuel Macron used the event to announce €109 billion in private investment commitments for AI infrastructure in France.

France’s gain also kept it well ahead of other large European economies on usage. Spain sits at No. 6, the U.K. at No. 8, Germany at No. 23 and Italy at No. 26.

Pierre Blaché / Unsplash

Norway ranks No. 5 in Microsoft $MSFT’s Global AI Diffusion report, with 49.4% of its working-age population using generative AI in June 2026. It fell from No. 3 in the first quarter, even though its usage rate went up.

Norway’s share rose from 48.6% to 49.4%, a gain of 0.8 percentage points. That was the smallest increase of any economy in the global top 10. Ireland and France both grew faster and moved past it. The three countries are now separated by only half a point, so the order could shift again in the next report.

Even after the drop in rank, Norway’s usage rate is more than two and a half times the global figure of 18.8%. Nearly one in every two Norwegians aged 15 to 64 used a generative AI product during the quarter.

Norway has long been one of the most digitally connected countries in the world. Internet access is close to universal, and Norwegians handle much of their banking, tax filing and public services online. BankID, a national electronic identity system used for logging into banks and government services, is a routine part of daily life.

The country is also wealthy and highly educated. Norway has about 5.6 million people and some of the highest incomes in Europe, supported by oil and gas revenue. That revenue flows into the Government Pension Fund Global, the world’s largest sovereign wealth fund. The fund holds stakes in thousands of listed companies worldwide.

What stands out in Microsoft’s data is how far Norway is ahead of its Nordic neighbors. Sweden ranks No. 17 at 37.3%, Denmark No. 22 at 32.4% and Finland No. 30 at 30.6%. Norway’s share is 12.1 points higher than Sweden’s and almost 19 points higher than Finland’s. That gap holds despite broad similarities in income, education and digital infrastructure across the region.

Norway also remains the top-ranked Nordic country and the third-highest in Europe, behind only Ireland and France.

Jorge Fernández Salas / Unsplash

Spain ranks No. 6 in Microsoft $MSFT’s Global AI Diffusion report, with 45.1% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter.

Spain’s share rose from 44.2% to 45.1%, a gain of 0.9 percentage points. The country sits 4.3 points behind Norway at No. 5 and one point ahead of New Zealand at No. 7. Its usage rate is about 2.4 times the global average of 18.8%.

With a population of about 49 million, Spain is one of the three most populous economies in the global top 10, along with France and the U.K. Madrid and Barcelona, its two largest cities, are the main centers of the country’s technology and startup activity.

Spain’s result is also well ahead of its Southern European neighbors. Italy ranks No. 26 at 31.4%, Portugal sits at 27.4% and Greece at 21.5%. Spain’s share is 13.7 points higher than Italy’s, even though the two countries have comparable incomes and close economic ties.

The Spanish government has moved early on AI oversight. In 2023, it set up the Spanish Agency for the Supervision of Artificial Intelligence in A Coruña in Galicia. The agency is known by its Spanish acronym, AESIA. It was the first national body in the European Union created specifically to supervise AI, ahead of the EU’s AI Act taking effect.

Spain also hosts major computing infrastructure. The Barcelona Supercomputing Center operates MareNostrum 5, one of the European Union’s most powerful supercomputers. Researchers use it for scientific modeling and AI training.

Language may be a factor in everyday use. Spanish is one of the world’s most widely spoken languages, with hundreds of millions of native speakers. After English, it is among the best-supported languages in major AI chatbots. That gives Spanish speakers access to more capable tools than speakers of many smaller languages.

Partha Narasimhan / Unsplash

New Zealand ranks No. 7 in Microsoft $MSFT’s Global AI Diffusion report, with 44.1% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter.

New Zealand’s share rose from 43.0% to 44.1%, a gain of 1.1 percentage points. That keeps it 3.4 points ahead of Australia, which ranks No. 11 at 40.7%. The two countries share many economic and cultural ties, but New Zealand has consistently posted the higher adoption rate in Microsoft’s tables.

New Zealand’s AI usage rate is about 2.3 times the global average of 18.8%. Its working-age population uses generative AI at a higher rate than those of the U.K. and the Netherlands. Its rate is also higher than that of any economy in East Asia.

The country is small, with about 5.3 million people, most of them concentrated in a few cities, led by Auckland. Its economy rests on agriculture, tourism and a large services sector, and it has high internet access and smartphone ownership. English is the dominant language of business and government. That gives most New Zealanders access to AI tools in the language where those tools have generally performed best.

New Zealand has a second official spoken language as well: te reo Māori. That matters for the future of AI use in the country. Most AI models are trained on far less text in Indigenous languages than in English.

Microsoft’s report points to work on that gap. Researchers at Microsoft’s AI for Good Lab built a framework called Bring Your Own Language. It adapts existing language models to under-resourced languages by refining training data. According to the report, the approach improved performance for Māori and Chichewa by about 12% on average across a set of benchmarks.

That kind of adaptation is one reason Microsoft’s report treats open-weight models as important to future diffusion. Developers can modify those models for local needs, including smaller languages, without training a system from scratch.

Charles Postiaux / Unsplash

The U.K. ranks No. 8 in Microsoft $MSFT’s Global AI Diffusion report, with 43.7% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter.

The U.K.’s share rose from 42.2% to 43.7%, a gain of 1.5 percentage points. That matched Ireland’s increase and was among the larger gains in the top 10. The U.K.’s usage rate is about 2.3 times the global average of 18.8%.

With roughly 69 million people, the U.K. is one of the two most populous economies in the global top 10, alongside France. Its adoption rate is 10.7 points higher than that of the U.S., which ranks No. 21. It is also 11.4 points higher than Germany’s, at No. 23.

London is one of the world’s main hubs for AI research. DeepMind was founded there in 2010 and acquired by Google $GOOGL in 2014. Now called Google DeepMind, it remains headquartered in King’s Cross and produced landmark systems such as AlphaGo and AlphaFold. Many other AI labs and startups have opened offices in London to tap the same talent pool. That pool draws heavily on universities such as Oxford, Cambridge, Imperial College London and University College London.

The British government has made AI a policy priority across administrations. In November 2023, it hosted the first global AI Safety Summit at Bletchley Park, the site of Britain’s World War II code-breaking effort. It then set up a state-backed institute to test advanced AI models, which was renamed the AI Security Institute in early 2025. In January 2025, the government published an AI Opportunities Action Plan aimed at expanding computing capacity and public-sector use of the technology.

The U.K.’s large financial and professional services sectors also matter. Banking, insurance, law, consulting and media employ a large share of the workforce in office roles. In those roles, AI tools for writing, research and analysis fit easily into daily work.

Tobias Kordt / Unsplash

The Netherlands ranks No. 9 in Microsoft $MSFT’s Global AI Diffusion report, with 43.5% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter.

The Dutch share rose from 42.1% to 43.5%, a gain of 1.4 percentage points. Qatar, at No. 10, posted the same 43.5% figure when rounded to one decimal place, but Microsoft lists the Netherlands ahead of it. The U.K. sits just 0.2 points higher at No. 8.

The Netherlands’ AI usage rate is about 2.3 times the global average of 18.8%. It is the highest-ranked economy among the Benelux countries, ahead of neighboring Belgium at No. 13.

The country plays an outsized role in the hardware behind AI. ASML $ASML, based in Veldhoven, is the only company that makes extreme ultraviolet lithography machines. Those systems are required to produce the most advanced chips, including the processors used to train and run large AI models. The surrounding Eindhoven region has built up a dense cluster of semiconductor suppliers and engineering firms.

The Netherlands is also a major node for global internet traffic. The Amsterdam Internet Exchange, known as AMS-IX, is one of the largest internet exchange points in the world. The region around Amsterdam hosts a large concentration of data centers serving European customers.

Language may help as well. The Netherlands consistently ranks at or near the top of the EF English Proficiency Index among countries where English is not a native language. High English fluency gives Dutch users easy access to AI tools that have generally performed best in English. Dutch-language support in major chatbots has also improved.

With about 18 million people, the Netherlands has a population larger than most of the economies ranked above it, apart from France, Spain and the U.K. Its economy is heavily weighted toward trade, logistics, finance and technology services. Office-based digital work is common in all of those sectors.

Visit Qatar / Unsplash

Qatar ranks No. 10 in Microsoft $MSFT’s Global AI Diffusion report, with 43.5% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter.

Qatar’s share rose from 41.8% to 43.5%, a gain of 1.7 percentage points. That was a faster increase than the Netherlands posted, and the two economies finished the quarter at the same rounded figure. Microsoft lists the Netherlands at No. 9 and Qatar at No. 10.

Qatar is the second-highest-ranked economy in the Middle East, behind only the UAE. The gap between the two Gulf neighbors is large, though. The UAE’s 73.3% share is nearly 30 points higher than Qatar’s. Qatar’s own rate is still about 2.3 times the global average of 18.8%.

Like the UAE, Qatar is a small, wealthy state with a population made up largely of foreign workers. About three million people live in the country, most of them in and around Doha. Natural gas exports have made Qatar one of the richest countries in the world per person. The government has spent heavily on digital infrastructure, education and research.

Much of that research spending runs through Qatar Foundation and its Hamad Bin Khalifa University. The university’s Qatar Computing Research Institute works on AI, Arabic-language computing and data science. Arabic has historically been less well supported in AI models than English, so local work on Arabic-language systems is a priority across the Gulf.

The country’s long-term development plan, Qatar National Vision 2030, aims to shift the economy away from heavy reliance on hydrocarbons toward knowledge-based industries. Expanding digital services and technology skills is part of that goal.

Qatar is one of three Gulf states in the global top 25 in Microsoft’s ranking. The UAE leads the table, and Saudi Arabia entered the top 25 in this quarter’s report at No. 25.

Caleb / Unsplash

Australia ranks No. 11 in Microsoft $MSFT’s Global AI Diffusion report, with 40.7% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter and crossed the 40% mark.

Australia’s share rose from 39.5% to 40.7%, a gain of 1.2 percentage points. It stays just ahead of South Korea, which climbed to No. 12 at 40.6%. Only 0.1 points separate the two economies, so Australia could easily lose its spot if South Korea’s rapid growth continues.

Australia also trails its closest neighbor. New Zealand ranks No. 7 at 44.1%, putting it 3.4 points ahead. Australia’s own rate is about 2.2 times the global average of 18.8%.

With about 27 million people, Australia is one of the larger economies in this part of the table. Its population is heavily urban and concentrated along the coast, mainly in Sydney, Melbourne, Brisbane, Perth and Adelaide. The country has high internet access and smartphone ownership, and English is its dominant language.

Australia has produced several major software companies. Atlassian $TEAM, founded in Sydney in 2002, makes workplace collaboration software used by large numbers of companies worldwide. Canva, the online design platform, was founded in Perth and is based in Sydney. Canva has added a range of generative AI features to its design tools, which puts AI in front of a large consumer and business user base.

The federal government has also built public capacity in the field. In 2021, it set up the National AI Centre, housed within CSIRO, the national science agency. The centre’s role is to help Australian businesses adopt AI responsibly.

Australia’s economy rests heavily on mining and agriculture exports, but most Australians work in services. Health care, education, finance and professional services employ large shares of the workforce. Those fields have been early places for generative AI tools to spread.

Mathew Schwartz / Unsplash

South Korea ranks No. 12 in Microsoft $MSFT’s Global AI Diffusion report, with 40.6% of its working-age population using generative AI in June 2026. It climbed four places from No. 16 and posted the largest gain of any economy in the quarter.

South Korea’s share rose from 37.1% to 40.6%, an increase of 3.5 percentage points. That was more than any other economy, including the UAE. According to Microsoft, it was the third straight reporting period in which South Korea recorded the largest gain of any major economy. Over the past 12 months, the country added nearly 15 points to its AI user share.

The jump moved South Korea past Belgium, Switzerland, Israel and Canada. It now sits just 0.1 points behind Australia at No. 11.

South Korea is one of the world’s most connected countries, with near-universal high-speed internet and very high smartphone ownership. Its population of about 51 million is heavily urban, centered on the Seoul metropolitan area.

The country is also central to the hardware that powers AI. Samsung Electronics and SK Hynix are the world’s two largest makers of memory chips. Both produce high-bandwidth memory, a specialized type of memory used alongside the processors that train and run large AI models. Demand for that memory has turned AI into a major part of South Korea’s export economy.

Korean companies have built their own AI models as well. Naver, the country’s largest search company, developed HyperCLOVA X $TWTR, a large language model designed with strong Korean-language ability.

South Korea has also moved quickly on regulation. In December 2024, its National Assembly passed the AI Basic Act, a broad framework law covering AI development, safety and trust. The law took effect in January 2026. That made South Korea one of the first countries after the European Union to adopt comprehensive AI legislation.

South Korea’s rise also widens its lead over Japan. Japan’s share reached 24.7% in the quarter, roughly 16 points below South Korea’s.

Jeffrey Zhang / Unsplash

Belgium ranks No. 13 in Microsoft $MSFT’s Global AI Diffusion report, with 40.3% of its working-age population using generative AI in June 2026. It slipped one place from No. 12, even though its usage rate rose.

Belgium’s share grew from 39.0% to 40.3%, a gain of 1.3 percentage points. That was enough to cross the 40% mark, but not enough to hold off South Korea, which jumped past it with a 3.5-point increase. Belgium now trails South Korea by just 0.3 points.

Belgium’s AI usage rate is a little more than twice the global average of 18.8%. Among its neighbors, it ranks behind the Netherlands at No. 9 and France at No. 4, but well ahead of Germany at No. 23.

Brussels sits at the center of AI regulation in Europe. The city hosts the main institutions of the European Union, where the EU’s Artificial Intelligence Act was drafted and negotiated. The law entered into force in August 2024, with its rules phasing in over several years. It sorts AI systems by risk level and sets obligations for providers of general-purpose models.

Belgium also has a significant role in chip research. Imec, based in Leuven, is one of the world’s leading research centers for semiconductor technology. Chipmakers and equipment suppliers from around the world work with imec on the manufacturing processes used for advanced processors, including those used in AI.

The country is small, with about 11.9 million people, and it is divided by language. Dutch is spoken mainly in Flanders in the north and French mainly in Wallonia in the south. German is spoken in a small eastern region. Brussels is officially bilingual in French and Dutch. Strong support for French and Dutch in major AI tools means most Belgians can use those tools in their first language.

Belgium’s economy leans heavily on services, including government, international organizations, finance, logistics and pharmaceuticals. Many of those jobs are office-based, which tends to favor quick uptake of AI writing and research tools.

Henrique Ferreira / Unsplash

Switzerland ranks No. 14 in Microsoft $MSFT’s Global AI Diffusion report, with 39.1% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter.

Switzerland’s share rose from 37.8% to 39.1%, a gain of 1.3 percentage points. It moved ahead of Israel, which was one place above it in the first quarter. South Korea’s jump past both economies left Switzerland’s rank unchanged.

The Swiss usage rate is a little more than twice the global average of 18.8%. Among German-speaking countries, Switzerland ranks highest by a clear margin. Austria sits at No. 18 with 35.4%, and Germany at No. 23 with 32.3%.

Switzerland’s universities are among the strongest in the world for computer science and engineering. ETH Zurich and EPFL in Lausanne train large numbers of AI researchers and run major research programs in the field. The Swiss National Supercomputing Centre in Lugano provides high-performance computing to Swiss researchers.

Those institutions have also put their weight behind open AI models. In 2025, ETH Zurich, EPFL and the Swiss National Supercomputing Centre released Apertus, a large language model. Its weights, training data details and code were made publicly available. That approach fits the trend Microsoft’s report describes, in which open-weight models are taking a growing share of AI usage among developers.

Large technology companies have a strong presence as well. Google $GOOGL runs one of its biggest engineering offices outside the U.S. in Zurich. Many international companies have research or regional headquarters in the country.

Switzerland has about nine million people and four national languages: German, French, Italian and Romansh. Most residents speak German, French or Italian, all of which are well supported in major AI tools. The country also has some of the highest incomes in the world. Its workforce is concentrated in finance, pharmaceuticals, precision manufacturing and international organizations, where digital office work is widespread.

Anton Lukin / Unsplash

Israel ranks No. 15 in Microsoft $MSFT’s Global AI Diffusion report, with 38.7% of its working-age population using generative AI in June 2026. It fell two places from No. 13 in the first quarter.

Israel’s share rose from 38.1% to 38.7%, a gain of 0.6 percentage points. That was the smallest quarterly increase of any economy in the global top 25. South Korea and Switzerland both grew faster and moved ahead of it.

Israel is the third-highest-ranked economy in the Middle East in Microsoft’s table, behind the UAE at No. 1 and Qatar at No. 10. It remains ahead of Saudi Arabia, which ranks No. 25. Israel’s usage rate is a little more than twice the global average of 18.8%.

The country has one of the world’s most developed technology sectors relative to its size. Israel has about 10 million people and a dense concentration of startups, venture capital, and research and development centers run by global companies. Intel $INTC has operated in Israel for decades and acquired Mobileye, the Jerusalem-based driver-assistance company, in 2017. Nvidia $NVDA bought Mellanox, an Israeli maker of high-speed networking equipment, in a deal completed in 2020. Nvidia’s Israeli operations now make networking technology used in AI data centers.

Israel’s universities and research institutes train large numbers of computer scientists and engineers. They include the Technion in Haifa, the Weizmann Institute of Science and Tel Aviv University. Many Israeli engineers gain early technical experience through military technology units before moving into the private sector.

Language may affect everyday adoption. Hebrew, the country’s main language, has far fewer speakers than English, Spanish or French, and less Hebrew text exists for training AI models. Major chatbots support Hebrew, but performance in smaller languages has tended to lag behind performance in English. Many Israelis working in technology use English professionally, which gives them easier access to the strongest tools.

Microsoft’s report does not explain why Israel’s growth slowed relative to the economies around it.

Adriana Macias / Unsplash

Canada ranks No. 16 in Microsoft $MSFT’s Global AI Diffusion report, with 38.5% of its working-age population using generative AI in June 2026. It slipped one place from No. 15 as South Korea moved past it.

Canada’s share rose from 37.3% to 38.5%, a gain of 1.2 percentage points. The country’s AI usage rate is about twice the global average of 18.8%. It is also 5.5 points higher than the rate in the U.S., which ranks No. 21 at 33.0%.

Canada has a deep history in AI research. In 2017, it became the first country to launch a national AI strategy, the Pan-Canadian Artificial Intelligence Strategy. That program funded three national AI institutes: the Vector Institute in Toronto, Mila in Montreal and Amii in Edmonton.

Several of the researchers whose work underpins modern AI built their careers at Canadian universities. Geoffrey Hinton spent decades at the University of Toronto, and Yoshua Bengio founded what became Mila at the Université de Montréal. Hinton and Bengio shared the 2018 Turing Award, computing’s highest honor, with Yann LeCun for their work on deep learning. Hinton later shared the 2024 Nobel Prize in Physics for foundational discoveries in machine learning.

Canada has commercial AI companies as well. Cohere, based in Toronto, builds large language models aimed mainly at businesses. Many U.S. technology companies run research offices in Toronto and Montreal.

Canada has about 41 million people, most of them living within a few hundred kilometers of the U.S. border. English and French are both official languages. Quebec, the largest French-speaking province, gives Canada a large population of French speakers. They can use AI tools that have strong French support.

The country’s high rate of urbanization, high internet access and large services economy also support adoption. Most Canadians work in fields such as health care, retail, finance, education and professional services, where generative AI tools have spread first.

Ana Bórquez / Unsplash

Sweden ranks No. 17 in Microsoft $MSFT’s Global AI Diffusion report, with 37.3% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter.

Sweden’s share rose from 36.1% to 37.3%, a gain of 1.2 percentage points. Its AI usage rate is about twice the global average of 18.8%.

Sweden is the second-highest-ranked Nordic country, behind Norway at No. 5. The gap between the two is wide. Norway’s share of 49.4% is 12.1 points above Sweden’s. Sweden in turn sits ahead of Denmark, at No. 22 with 32.4%, and Finland, at No. 30 with 30.6%.

The country has a strong track record in consumer technology. Spotify $SPOT, the music streaming service, was founded in Stockholm in 2006 and remains headquartered there. Klarna, the payments company, also started in Stockholm. It has publicly described using AI to handle a large share of its customer service work. Ericsson, one of the world’s largest makers of mobile network equipment, is based in the Stockholm area as well.

Sweden has also invested heavily in AI research through private funding. The Wallenberg AI, Autonomous Systems and Software Program, known as WASP, was launched in 2015 with funding from the Knut and Alice Wallenberg Foundation. It has become one of the largest research programs in Sweden’s history, supporting doctoral students, faculty hires and research at several universities.

Sweden has about 10.6 million people and very high internet access. Swedes do much of their banking, shopping and government business online, and cash has largely disappeared from everyday transactions. English proficiency is also high, giving Swedish users easy access to AI tools that have generally performed best in English.

Sweden’s place in the table shows that strong digital infrastructure does not translate into identical adoption rates. Norway, Sweden, Denmark and Finland are broadly similar in income and connectivity. Yet the four countries are spread across a range of nearly 19 points in Microsoft’s data.

Leonhard Niederwimmer / Unsplash

Austria ranks No. 18 in Microsoft $MSFT’s Global AI Diffusion report, with 35.4% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter.

Austria’s share rose from 34.1% to 35.4%, a gain of 1.3 percentage points. Its AI usage rate is a little under twice the global average of 18.8%.

Austria sits between its two largest German-speaking neighbors in the table. Switzerland ranks four places higher at No. 14, with a share 3.7 points above Austria’s. Germany ranks five places lower at No. 23, with a share 3.1 points below. Austria is also ahead of Hungary, its eastern neighbor, which ranks No. 20.

Austria has an important place in the history of modern AI. Sepp Hochreiter, now a professor at Johannes Kepler University Linz, co-invented the long short-term memory network with Jürgen Schmidhuber in the 1990s. That network is known as LSTM. LSTM networks were widely used for speech recognition, translation and text prediction in the years before the transformer architecture became dominant. Hochreiter has continued that work in Linz with a newer architecture called xLSTM.

Vienna is the country’s main technology and research center. The Institute of Science and Technology Austria, in Klosterneuburg just outside the city, is a graduate research institute with programs in computer science and machine learning. TU Wien and the University of Vienna also run large computer science departments.

Austria has about 9.2 million people, roughly a fifth of whom live in Vienna. German is the official language nationwide and is among the best-supported languages in major AI tools. The country has high incomes, high internet access and an economy built around manufacturing, tourism, finance and services.

Austria’s position in the table reflects a steady, middle-of-the-pack pace. Its quarterly gain matched that of Switzerland, Belgium and Poland, and was slightly larger than Germany’s. That steady growth kept its rank unchanged at No. 18.

Vas / Unsplash

Taiwan ranks No. 19 in Microsoft $MSFT’s Global AI Diffusion report, with 33.8% of its working-age population using generative AI in June 2026. It moved up one place from No. 20, passing Hungary.

Taiwan’s share rose from 31.8% to 33.8%, a gain of 2.0 percentage points. Only South Korea and the UAE grew faster among the top 25, and Saudi Arabia matched Taiwan’s increase. Taiwan’s AI usage rate is about 1.8 times the global average of 18.8%.

Microsoft’s report ranks “economies” rather than countries, which is how Taiwan appears in the table alongside sovereign states.

Taiwan is central to the global AI supply chain. Taiwan Semiconductor Manufacturing Company, known as TSMC $TSM, manufactures the most advanced chips in the world. Those include the graphics processors that Nvidia $NVDA designs for training and running AI models. Foxconn, formally known as Hon Hai Precision Industry, assembles a large share of the AI servers built around those chips.

That role in hardware has not made Taiwan one of the very top users of AI. Its usage rate trails those of Singapore and South Korea, the two highest-ranked economies in East and Southeast Asia in Microsoft’s data. The 2.0-point gain this quarter suggests that gap may be narrowing.

Taiwan is ahead of other large East Asian economies by a wide margin. Japan reached 24.7% in the quarter, 9.1 points below Taiwan. China stands at 17.5% in Microsoft’s current data. The report says its next edition will add newer AI tools that it does not yet measure. It expects that change to raise usage figures almost everywhere, with the largest increase in China. Several widely used Chinese domestic tools are not included in the current measure.

Taiwan has about 23 million people and very high internet and smartphone use. Mandarin Chinese is the main language, written in traditional characters. Its economy is built around electronics manufacturing, chip design and exports.

Ervin Lukacs / Unsplash

Hungary ranks No. 20 in Microsoft $MSFT’s Global AI Diffusion report, with 33.3% of its working-age population using generative AI in June 2026. It slipped one place from No. 19 as Taiwan moved ahead of it.

Hungary’s share rose from 32.2% to 33.3%, a gain of 1.1 percentage points. Its AI usage rate is about 1.8 times the global average of 18.8%.

Hungary is the highest-ranked economy in Central and Eastern Europe in Microsoft’s table. Poland follows at No. 24 with 32.3%, the Czech Republic at No. 27 with 31.2% and Bulgaria at No. 29 with 30.7%. Hungary also ranks above two much larger and richer economies: the U.S., at No. 21, and Germany, at No. 23.

Hungary has about 9.6 million people, with roughly a fifth of them living in Budapest, the capital. The city is the country’s main center for software development, business services and university research.

Hungarian is an unusual language for AI. It belongs to the Uralic language family, related distantly to Finnish and Estonian. Most other European languages, including English and German, belong to the Indo-European family instead. Hungarian has relatively few speakers worldwide, so far less Hungarian text exists for training AI models. Even so, Hungarian adoption is higher than in many economies whose languages are much better represented in AI training data.

The country’s economy has a large manufacturing base, especially in the auto industry. Audi, Mercedes-Benz and BMW all operate or are building plants in Hungary. Battery makers have invested in factories to supply electric vehicle production. Alongside that industry, shared service centers run by multinational companies employ many Hungarians in finance, IT and customer support roles. AI tools are increasingly part of daily work in those roles.

Hungary’s slip from No. 19 to No. 20 came from Taiwan’s faster growth, not from a decline. Its own usage continued to rise, keeping it ahead of most of its European Union neighbors to the east.

Michael Discenza / Unsplash

The U.S. ranks No. 21 in Microsoft $MSFT’s Global AI Diffusion report, with 33.0% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter.

The U.S. share rose from 31.3% to 33.0%, a gain of 1.7 percentage points. That was one of the larger increases in the top 25, but it did not change the country’s rank. The U.S. usage rate is about 1.75 times the global average of 18.8%.

The ranking contrasts with the U.S. role in building the technology. OpenAI, Anthropic, Google $GOOGL, Microsoft, Meta $META and xAI, the companies behind most of the world’s leading AI models, are all American. Nvidia $NVDA, the dominant maker of AI chips, is headquartered in California. U.S. companies and investors account for the largest share of private investment in AI worldwide.

Microsoft’s measure is not designed to capture any of that. It tracks only the share of working-age people who used a generative AI product during the period. It does not measure how intensively those users work with AI, how much businesses spend on it, or where AI models are built.

On that narrower metric, the U.S. trails many peer economies by a wide margin. The U.K.’s rate is 10.7 points higher, at 43.7%. Canada’s is 5.5 points higher, at 38.5%. Australia, New Zealand, France, Ireland and the Netherlands all rank well ahead of the U.S. as well.

The U.S. has about 340 million people, by far the largest population of any economy in the top 25. A national rate reflects a very wide range of income levels, ages, industries and internet access across states and regions. Microsoft’s report does not break down U.S. results below the national level.

The U.S. sits in a tight cluster with several European economies. Hungary, one place above at No. 20, is just 0.3 points higher. Denmark, Germany and Poland, ranked No. 22 through No. 24, are all within 0.7 points below.

Bob Jenkin / Unsplash

Denmark ranks No. 22 in Microsoft $MSFT’s Global AI Diffusion report, with 32.4% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter.

Denmark’s share rose from 31.2% to 32.4%, a gain of 1.2 percentage points. Its AI usage rate is about 1.7 times the global average of 18.8%. It sits 0.6 points behind the U.S. and 0.1 points ahead of Germany and Poland.

The result places Denmark well below its Nordic neighbors at the top of the table. Norway ranks No. 5 at 49.4%, a rate 17 points higher than Denmark’s. Sweden ranks No. 17 at 37.3%. Among the Nordic countries in Microsoft’s top 30, only Finland, at No. 30, ranks lower.

That gap stands out because Denmark is one of the most digitized countries in the world. It ranked first in the United Nations E-Government Survey in 2024, as it did in several earlier editions. Danes use a national digital ID, MitID, to access banking and public services. Most communication with the government happens online through a digital mailbox.

Denmark has also invested in AI computing power. In 2024, the Novo Nordisk Foundation and Denmark’s Export and Investment Fund opened Gefion, a national AI supercomputer built with Nvidia $NVDA technology. It is housed at a new Danish center for AI innovation. The system is intended for researchers and companies working in fields such as drug discovery, quantum computing and climate science.

Novo Nordisk, the maker of the weight-loss drug Wegovy and the diabetes drug Ozempic, is Denmark’s largest company. Its foundation is one of the world’s largest charitable foundations and funds a wide range of Danish science.

Denmark has about six million people and very high incomes. Danish is spoken by relatively few people worldwide, but English proficiency in the country is high. That combination means most Danes can use AI tools in English even where Danish-language performance lags behind.

Florian Wehde / Unsplash

Germany ranks No. 23 in Microsoft $MSFT’s Global AI Diffusion report, with 32.3% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter.

Germany’s share rose from 31.1% to 32.3%, a gain of 1.2 percentage points. Its AI usage rate is about 1.7 times the global average of 18.8%. Poland, at No. 24, posted the same 32.3% figure when rounded to one decimal place, with Microsoft listing Germany ahead.

Germany is Europe’s largest economy and its most populous country, with about 84 million people. Yet it trails most of its neighbors in Microsoft’s table. France ranks No. 4, the Netherlands No. 9, Belgium No. 13, Switzerland No. 14, Austria No. 18 and Denmark No. 22. Ireland’s 49.9% share is 17.6 points higher than Germany’s.

Germany has built a set of homegrown AI companies. DeepL, based in Cologne, makes machine translation software used by businesses and individuals around the world. Aleph Alpha, based in Heidelberg, develops AI systems aimed at government and corporate customers. Black Forest Labs, founded in Freiburg in 2024, builds image generation models.

The country also has a long research tradition in the field. The German Research Center for Artificial Intelligence, known by its German acronym DFKI, was founded in 1988. It is one of the largest nonprofit AI research centers in the world. Universities such as the Technical University of Munich and the University of Tübingen run major machine learning programs.

Germany’s economy is built heavily around manufacturing. Automakers, chemical companies, machinery makers and the thousands of mid-sized family firms known as the Mittelstand employ a large share of the workforce. Many of those jobs are on factory floors or in technical roles. Consumer generative AI tools play a smaller part in daily work there than they do in office settings.

German is among the best-supported languages in major AI tools, so language access is unlikely to be a major barrier for most German users.

Paweł Malinowski / Unsplash

Poland ranks No. 24 in Microsoft $MSFT’s Global AI Diffusion report, with 32.3% of its working-age population using generative AI in June 2026. It held the same position it occupied in the first quarter.

Poland’s share rose from 31.0% to 32.3%, a gain of 1.3 percentage points. That matched Germany’s figure for the quarter when rounded to one decimal place, though Microsoft lists Germany one place higher. Poland’s AI usage rate is about 1.7 times the global average of 18.8%.

Poland is the second-highest-ranked economy in Central and Eastern Europe in Microsoft’s table, behind Hungary at No. 20. It ranks ahead of the Czech Republic at No. 27 and Bulgaria at No. 29.

With about 37 million people, Poland is the largest economy in Central Europe. It has grown quickly since joining the European Union in 2004. It has become a major location for software development, IT services and business process work for companies across Europe and the U.S.

Polish engineers have played a visible role in the AI industry. Wojciech Zaremba, a co-founder of OpenAI, and Jakub Pachocki, OpenAI’s chief scientist, both studied in Poland. Many other Polish computer scientists work at major AI labs abroad and at research teams in Warsaw, Kraków and Wrocław.

Poland’s technology sector also includes globally known game developers. CD Projekt, based in Warsaw, created The Witcher and Cyberpunk 2077 video game series. A large pool of programmers trained in the country’s technical universities supports both gaming and IT services.

Polish is a Slavic language with around 40 million speakers, far fewer than English or Spanish. Polish researchers have worked on language models built specifically for Polish to improve performance in the language. That focus on local-language AI reflects a concern Microsoft’s report raises more broadly: models adapted to local languages could make AI more useful for people outside the English-speaking world.

Oualid Lakhrouti / Unsplash

Saudi Arabia ranks No. 25 in Microsoft $MSFT’s Global AI Diffusion report, with 31.4% of its working-age population using generative AI in June 2026. It climbed five places from No. 30, the largest rise in rank of any economy in the quarter.

Saudi Arabia’s share rose from 29.4% to 31.4%, a gain of 2.0 percentage points. That tied Taiwan for the third-largest increase among the top 25, behind South Korea and the UAE. Italy posted the same 31.4% figure, but Microsoft lists Saudi Arabia ahead of it at No. 25, with Italy at No. 26.

The climb moved Saudi Arabia past Italy, the Czech Republic, Jordan, Bulgaria and Finland, and into the top 25. Its AI usage rate is now about 1.7 times the global average of 18.8%.

Saudi Arabia is one of three Gulf states in the global top 25, along with the UAE at No. 1 and Qatar at No. 10. It still trails its neighbors by a wide margin. The UAE’s share is more than twice as high, and Qatar’s is 12.1 points higher.

The kingdom has made technology central to Vision 2030, the economic plan launched in 2016 to reduce its reliance on oil. In 2019, it set up the Saudi Data and Artificial Intelligence Authority, known as SDAIA, to lead national policy on data and AI. In 2025, the Public Investment Fund, the country’s sovereign wealth fund, launched Humain. The company focuses on AI models, data centers and AI services, including Arabic-language AI.

Saudi Arabia has about 35 million people, making it by far the most populous Gulf state. Its population is young, with a large share of citizens under 30, and smartphone and internet use are very high. Arabic is the official language, and improving Arabic-language AI has been a stated priority of the country’s AI programs.

Other Gulf states rank lower in Microsoft’s data. Oman stands at 27.6% and Kuwait at 22.6%, both below Saudi Arabia’s new figure.