Published on

22 Ways to Understand AI: Europe's AI Literacy Education Revolution

"Knowing how to use AI" and "understanding AI" are two very different things. Just as billions of people use smartphones without understanding semiconductors, using an AI tool is entirely different from grasping what AI is, how it works, and what its limitations and risks actually are.

The European Union and the OECD have concluded that this distinction will be critically important for the next generation. In May 2025, the two organizations jointly released a draft of an AI Literacy Framework for primary and secondary education β€” with the final version due in 2026. Structured around 22 core competencies that young people need to navigate an AI-shaped world, this framework may reshape the landscape of European education.

Table of Contents

  1. Why AI Literacy? The Question 77% Can't Answer
  2. The Framework: 4 Domains, 22 Competencies
  3. The EU AI Act Connection: Literacy Becomes a Legal Requirement
  4. What's Already Happening in Classrooms
  5. PISA 2029: AI Literacy Goes on the Test

1. Why AI Literacy? The Question 77% Can't Answer

The numbers are striking. According to an EY survey, 77% of European companies report difficulty finding talent with AI skills. AI is already a fundamental tool in the workplace β€” yet the people who can work with it meaningfully are in critically short supply.

More revealing still is how students see their own situation. In a 2024 survey of 7,000 young people aged 12–17, 74% said they expect AI to significantly influence their future careers. Yet only 46% felt their schools were currently preparing them for this reality.

Students already know. Their schools are not keeping up.

2. The Framework: 4 Domains, 22 Competencies

The OECD–EU joint AI Literacy Framework is organized into four domains. Each moves beyond "learning to use a tool" toward a more rounded, critical relationship with AI.

Domain 1: Engage with AI

Understanding what AI is and how it works at a foundational level β€” and being able to use AI tools appropriately. This is not about memorizing how to operate software; it includes understanding the principles behind how AI generates its outputs.

Domain 2: Create with AI

The capacity to use AI as a creative partner β€” for writing, image generation, coding, and more. Crucially, this domain asks students to think about what role human judgment and creativity must play when AI is involved in making something.

Domain 3: Manage AI

The ability to evaluate AI critically and use it responsibly. This means recognizing AI bias, understanding the possibility of errors and hallucinations, being alert to privacy risks, and knowing when to trust AI β€” and when to question it.

Domain 4: Design AI

Understanding how AI systems are built and what values they should reflect. This does not mean every student must learn to develop AI. But by understanding how AI is trained, what data it learns from, and how it makes decisions, students can engage meaningfully with the social consequences of AI β€” rather than simply being subject to them.

The 22 specific competencies drawn from these four domains span technical understanding, critical thinking, ethical judgment, and humanistic perspective. This is not just about coding skills. It includes recognizing algorithmic bias, understanding the limits of AI, and thinking philosophically about the relationship between human creativity and machine output.

3. The EU AI Act Connection: Literacy Becomes a Legal Requirement

This framework carries more weight than a policy recommendation β€” because EU law is already pushing in the same direction.

On February 2, 2025, Article 4 of the EU AI Act entered into force, establishing an obligation for providers and deployers of AI systems to ensure that their staff have a sufficient level of AI literacy β€” taking into account their technical knowledge, experience, education, and training.

This is not only a corporate concern. Teachers, school administrators, and education policymakers are all increasingly engaged in AI-related decision-making. The legal pressure to understand AI is real, and it is beginning to filter into schools.

The EU's Union of Skills initiative set out a 2030 Roadmap for digital education, explicitly naming AI literacy for primary and secondary students as a core objective β€” alongside teacher support and the ethical integration of AI into educational institutions.

4. What's Already Happening in Classrooms

Change is not waiting for frameworks to be finalized.

The AI-ENTR4YOUTH program β€” coordinated by JA Europe, with support from Intel and the European Commission β€” is already running in 10 European countries, guiding secondary school students through applied AI projects. Beginning in Italy, Portugal, and Spain, it has expanded to Albania, Bulgaria, Czech Republic, France, Greece, Romania, and Ukraine.

The curriculum covers data literacy, AI ethics, no-code AI tools, computer vision, Python programming, mathematics for AI, and business model development β€” right through to pitching and prototyping. It is an integrated education that combines technical skills with social and economic context.

5. PISA 2029: AI Literacy Goes on the Test

The true measure of any education framework is whether it is actually assessed. The OECD plans to include Media and AI Literacy as a new domain in the PISA 2029 international assessment.

The implications are significant. Like mathematics, science, and reading, the ability to understand and critically engage with AI will become an official international measure of educational achievement. Curricula around the world will likely be reshaped accordingly.

Korea, Japan, the United States, Europe β€” no country is exempt from this wave. And at its center is a vision of education that goes beyond teaching students to "use" AI, toward helping them grow into people who can understand, question, and shape it.

The heart of AI education is not coding ability. It is the capacity to think about how AI affects human life and society β€” and the will to engage with that impact responsibly.

Of the 22 competencies, perhaps the most important is this: the ability to ask questions about AI.

Further Reading

Sources

Discuss on Twitter β€’ View on GitHub
22 Ways to Understand AI: Europe's AI Literacy Education Revolution | MINSSAM.COM