Artificial intelligence (AI) is becoming part of daily life. The tools and opportunities offered by major AI companies have grown rapidly in recent years. Education is no exception. Authorities, teachers or scholars reveal how the students use tools such as ChatGPT to complete homework, including essays. At the same time, AI also offers valuable support for teachers.
How widely are teachers using AI across Europe? Which countries are ahead? And what do teachers rely on AI for the most?
According to the Organisation for Economic Co-operation and Development’s (OECD) Teaching and Learning International Survey (TALIS), teachers’ use of AI significantly varies across Europe as of 2024.
Among 32 countries, the share of lower secondary school teachers using AI ranges from 14 per cent in France to 52 per cent in Albania as of 2024. This is 32 per cent on average in the European Union (EU-22), while it is 36 per cent in the OECD (27 countries).
In the survey, AI use includes making predictions, suggesting decisions or generating text. It covers the use of AI in teaching or to facilitate student learning in the 12 months prior to the survey.
Strong regional patterns do not clearly emerge, although in general, Western European countries tend to have lower teacher use of AI compared with the Western Balkans and Eastern Europe.
Besides Albania, the share of teachers who used AI at least once reached two in five or more in eight countries/economies. Their shares in percentages were Malta (46), Czechia (46), Romania (46), Poland (45), Kosovo (43), North Macedonia (42), Norway (40) and the Flemish region of Belgium (40).
The countries that used AI in schools the least were in Bulgaria (22), Hungary (23), the French region of Belgium (23), Turkey (24), Italy (25), Finland (27), Montenegro (28), and Slovakia (29).
Why does AI use widely vary?
A spokesperson from the United Nations Educational, Scientific and Cultural Organization (UNESCO) noted that governments have adopted varying policy positions on AI in education, which may have a bearing on teacher awareness and use of AI in Europe.
“Some countries have been more proactive in driving national AI strategies, which include the education sector, while others have taken a cautionary stance on AI and the use of generative AI in classrooms, providing stricter rules depending on the age of students,” the spokesperson told Euronews Next.
Ruochen Li, senior project manager at OECD, noted that “infrastructure, technological restrictions such as firewalls, social attitudes towards the use of technology in schools and policies that might encourage or discourage teachers from using AI” could explain the disparities across the countries.
“We do see a strong relationship at the country-level between the amount of training offered about using AI and the use of AI,” he told Euronews Next.
Ben Hertz and Antoine Bilgin, pedagogical & research managers at European Schoolnet, pointed out that the wide differences in AI use reflect each country’s policy environment and educational culture, with some taking a more precautionary national stance.
“Access to practical support like training and infrastructure is also a critical accelerator. TALIS data confirms that in systems with higher AI use, teachers are more likely to have received professional learning on the topic,” Hertz and Bilgin told Euronews Next.
For example, France began rolling out nationwide AI training in public schools this year, which came after the TALIS data was collected.
“Caution, unclear rules, and limited infrastructure are likely factors that explain slower uptake, especially in a new and controversial area like AI,” they added.
Possible reasons for this gap are the presence and quality of training, workload, teachers’ shortage, personal motivation and curiosity, according to Martina Di Ridolfo, policy coordinator at the European Trade Union Committee for Education (ETUCE).
What do teachers use AI for?
Out of teachers who used AI, on average in the European Union (EU-22), almost two-thirds (65 per cent) say that they use it to efficiently learn about and summarise a topic, and 64 per cent use AI to generate lesson plans or activities. The two highest percentages indicate that AI is used mainly for teachers’ own preparation.
Other purposes and their shares are:
- Help students practice new skills in real-life scenarios (49)
- Support students with special education needs (40)
- Automatically adjust the difficulty of lesson materials according to students’ learning needs (39)
- Generate text for student feedback or parent/guardian communications (31)
- Review data on student participation or performance (29)
- Assess or mark student work (26)
These findings suggest that direct classroom or student-facing uses are less common. Teachers mainly rely on AI for their own preparation, while assessment tasks are the least used. Hertz and Bilgin suggested that many teachers are likely to use AI mainly ‘behind the scenes’ for now and in the near future.
OECD’s Ruochen suggested that AI can support teachers with administrative work, and this could free up their time and energy for other tasks more closely related to direct instruction.
What’s next for AI in schools?
Experts agree that AI use in education is steadily growing and will continue to expand. They also caution that responsible use, clear guidelines, and awareness of possible drawbacks must be part of this progress.
Hertz and Bilgin of European Schoolnet noted that over time, AI systems will possibly interact more often directly with students, for example, by proposing personalised activities or providing real-time feedback.
“But teachers should remain the key interlocutor between the student and the technology to preserve their agency, ethical oversight and care role,” they said.
UNESCO emphasises the central role of teachers as AI use in education expands.
“AI tools must complement, not replace, teachers, and that their use should align with ethical standards and educational goals, preserving teachers’ and learners’ autonomy and privacy,” UNESCO spokesperson said.
Expecting AI use to grow among both teachers and students, especially with generative tools, Di Ridolfo highlighted another concern: “Given the severe teacher shortage we are facing in Europe, we see concrete risks of de-qualification and de-skilling of the profession,” she told Euronews Next.
She also drew attention to a limitation in the OECD survey: Data doesn’t say anything about the frequency of this use. It doesn’t say whether teachers have used AI regularly or only tested a few times.