Which Countries' Teachers Actually Use AI? The OECD Just Answered That
Everyone has an opinion about AI in classrooms. What's been missing, until recently, was real data comparing how teachers actually use it — not survey hype, not vendor marketing, but a standardized measurement across dozens of countries asking the exact same questions.
That's what the OECD's Teaching and Learning International Survey (TALIS) 2024 delivered. Nearly 280,000 teachers across 55 education systems were asked, among many other things: do you use AI in your work, and if so, how? The results, released in October 2025, are the first real global benchmark of AI adoption in teaching — and the gap between the top and bottom of the list is bigger than most people expect.
The Leaders: Singapore and the UAE, by a Wide Margin
Two education systems stand out clearly above everyone else: Singapore and the United Arab Emirates, where roughly 75% of teachers report using AI in their work — more than double the OECD average of 41%.
Singapore's number isn't an accident. The country has built AI tools directly into its national Student Learning Space platform, giving teachers built-in access to lesson-plan generation, student needs analysis, and marking assistance without needing to seek out commercial tools themselves. The results show up in teacher attitudes too: 82% of Singaporean teachers say AI helps them build or improve lesson plans, well above the OECD average of 53%. Singapore's teachers also lead the world in hybrid and online teaching experience, with over 80% working in schools offering blended lesson formats — more than five times the OECD average — suggesting the AI adoption isn't happening in isolation. It's part of a broader digital-first teaching culture the country built up systematically, not just a recent AI trend layered on top.
The UAE tells a similar story from a different angle. Vietnam — not one of the two AI-adoption leaders, but notable in its own right — posted the single highest agreement rate in the entire survey on whether AI genuinely helps lesson planning: 91% of Vietnamese teachers said yes, even higher than the UAE's already strong 87%.
The Laggards: France and Japan
At the other end of the ranking sit France and Japan, where fewer than 20% of teachers — in France's case, as low as 14% — report using AI at all. That's not a small gap. It means a French teacher is roughly five times less likely to be using AI in their work than a colleague in Singapore or the UAE.
France's low number comes with an attitude gap too: only 18% of French teachers agree that AI actually helps improve their lesson plans, compared to that 87-91% range in the UAE and Vietnam. That's not simply about access to tools — it points to a much deeper skepticism about whether AI is worth the effort at all, a pattern echoed in several other European systems where adoption also lags the global leaders.
The Bigger Pattern Behind the Numbers
Two things stand out once you look past the individual country rankings.
First, infrastructure and training explain a lot, but not everything. Among teachers worldwide who reported not using AI, 75% said it was because they lacked the knowledge or skills to use it — not because they were opposed to it in principle. That's a training-and-access problem, not a resistance problem, and it means the gap between Singapore and France is at least partly closeable with the right investment.
Second, usage doesn't automatically mean confidence. Even in high-adoption systems, TALIS found most teachers worldwide still feel unprepared to use AI responsibly in the classroom — meaning the "how much" question the survey answered clearly is a very different question from "how well," which remains far murkier even in the countries topping the adoption charts.
Where Does That Leave a System Like Azerbaijan's?
Azerbaijan wasn't part of the 55 systems surveyed in TALIS 2024, so there's no official number to point to directly. But the pattern the survey reveals is instructive regardless: the countries leading in AI adoption didn't get there through enthusiasm alone. Singapore built AI directly into a national teaching platform; the UAE and Vietnam paired strong adoption with strong teacher belief that the tools actually help. The lagging systems, meanwhile, aren't lagging because their teachers are less capable — they're lagging because the training and infrastructure gap hasn't been closed yet.
For a country like Azerbaijan, already investing in digital-skills programs and holding direct meetings with organizations like OpenAI, the TALIS data offers a fairly clear roadmap: adoption follows infrastructure and training, not the other way around. The countries at the top of this list didn't wait for teachers to figure AI out on their own.
