The European Commission and OECD propose that students should not only use artificial intelligence but also critically engage with it.

The European Commission and OECD have presented a framework for artificial intelligence literacy for primary and secondary education, at a time when students are already massively using AI tools for homework, research, translation, creation, or online interactions. The document proposes that schools should teach children not only to use AI applications but also to understand how they work, when they are wrong, how they can be manipulated, what data they use, and how they can be used responsibly.

The European Commission and OECD have presented a framework for artificial intelligence literacy that shifts the discussion in schools beyond the question of whether students are allowed to use ChatGPT or other similar tools. The stakes of the document are broader: children need to learn how AI works, how to verify generated responses, when it is appropriate to use it, what risks exist for critical thinking, privacy, creativity, emotional health, and democracy.

In short The European Commission and OECD have launched a framework for AI literacy for primary and secondary education. The document is non-binding and addresses teachers, school leaders, educational authorities, training providers, and families. The framework is organized around four areas: interaction with AI, creation with AI, managing AI, and modeling AI. The report shows that AI literacy does not mean simply using tools, but understanding how AI works and critically assessing its role in everyday life. The framework will contribute to the PISA 2029 assessment on media literacy and artificial intelligence.

The document, titled "Empowering Learners for the Age of AI," starts from a reality already visible in schools: students use artificial intelligence before educational systems have clearly established what they need to understand about it. AI appears in chatbots, translation applications, recommendation engines, search tools, creative applications, educational platforms, and services that influence what information young people see.

The framework defines AI literacy as a set of technical knowledge, sustainable skills, and attitudes necessary to live in a world influenced by artificial intelligence. In simple terms, students should be able to interact with AI, create with its help, decide when and how to use it, and understand how it can be improved to reflect human values.

This distinction is important because the report insists that AI literacy is not the same as using AI tools. A student can enter a request into a chatbot without understanding why the response may be wrong, biased, fabricated, or manipulative. The framework proposes that schools should develop this capacity for judgment, not just the technical ability to write prompts.

The document also indicates the scale of the phenomenon. A study cited in the framework shows that 88% of younger European teenagers, aged 13 to 15, and 96% of teenagers aged 16 to 18 use AI tools for learning and creative tasks at least a few times a week. Eurostat also shows that young people aged 16 to 24 use generative AI almost twice as much as the general population.

For schools, the issue is no longer whether students will encounter AI, but whether they will use it judiciously. The report mentions risks related to misinformation, inappropriate content, algorithmic bias, gender stereotypes, privacy, digital footprints, and the use of AI companions as substitutes for human interactions.

An important chapter addresses the confusion between AI and humans. The framework explains that AI systems can produce responses that seem like human conversations but lack consciousness, understanding, intention, or authentic relationships. In the case of young people, this confusion can become sensitive when chatbots or AI companions are used for emotional support, serious conversations, or sharing personal information.

The framework is built around four areas. The first, "interaction with AI," aims for students to recognize where artificial intelligence appears in their lives, evaluate the generated outcomes, and decide whether these should be accepted, revised, or rejected. This includes information verification, identifying hallucinations, understanding algorithmic recommendations, and analyzing the impact of AI on opinions, choices, and behaviors.

The second area, "creation with AI," concerns using artificial intelligence as a creative partner without losing human control. Students should be able to use AI to explore ideas, visualize or prototype solutions, seek feedback, and reflect on originality, copyright, and authenticity of content.

The third area, "managing AI," encourages students to decide whether AI is the right tool for a specific task. The framework proposes that young people compare what an AI system can do with what should remain human responsibility, such as judgment, creativity, relationships, reasoning, or understanding context.

The fourth area, "modeling AI," goes beyond using existing tools. Students should understand how AI systems are designed, for whom they are built, what limitations they have, what data they use, and how they can be improved to promote human well-being and social benefit. The report specifies that the goal is not to turn all students into AI engineers but to train young people who know that technology can be evaluated, adjusted, and contested.

Teachers are presented as central actors, but not as the only ones responsible. The framework asserts that AI literacy must be built through a coordinated effort among teachers, school leaders, educational authorities, training providers, parents, and families. The message is that a single teacher or a single subject cannot cover the entire issue.

The data cited in the report also show the difficulty of this transition. A 2025 Eurobarometer Flash indicates that 81% of European citizens believe that all teachers should have the necessary skills to use and understand AI, including generative AI. At the same time, OECD's TALIS 2024 data show that only one in three teachers uses AI on average, and three out of four teachers say they do not have the knowledge and skills necessary to teach using AI.

The report warns that introducing AI in schools should not be treated as a simple time-saving method. Teachers remain essential for empathy, intellectual curiosity, resilience, ethical reasoning, inclusion, and individualized support. The framework insists that technology should serve the development of students, not undermine the professional autonomy of teachers or the human relationship in education.

The document also discusses the risks for learning. AI can help students produce better work or reach certain outcomes faster, but the report warns that these benefits do not automatically translate into sustainable learning. Delegating cognitive effort to AI can weaken critical thinking and self-regulation, especially if students accept responses without verification or reflection.

Ethics is not treated as a separate chapter but as a dimension present throughout the framework. Students need to understand that AI systems reflect human choices, selected data, design criteria, economic interests, working conditions, and social inequalities. The framework emphasizes transparency, accountability, confidentiality, fairness, environmental impact, and the central question: who benefits and who may be disadvantaged by the use of AI.

A less discussed element in the public debate, but explicitly included in the framework, is the environmental cost of artificial intelligence. Students should learn that AI systems use energy, water, minerals, and computing infrastructure, and the decision to use AI must be weighed against its real relevance for the respective task.

The framework was developed after a preliminary version launched in May 2025 and revised following an international consultation. Over 2,000 people contributed feedback from more than 100 countries, including teachers, learning designers, decision-makers, researchers, NGOs, unions, parents, and youth representatives.

Roxana Mînzatu, Executive Vice-President of the European Commission for social rights and skills, quality jobs, and training, stated that AI literacy is "a fundamental skill" for young people. She asserted that understanding and responsibly using AI are essential for Europe's competitiveness, but also for protecting democracy and social cohesion, and these things "start in our classrooms."

The framework is non-binding and does not provide guidance for the legal application of the European regulation on artificial intelligence. Its role is educational: to provide a common language and concrete examples for schools, so that AI literacy can be introduced into different subjects and adapted to national, local, and school contexts.

The document will contribute to the PISA 2029 assessment on media literacy and artificial intelligence, which will measure students' ability to critically navigate an information environment increasingly influenced by AI systems. For the European Union, the framework is linked to the Digital Education Action Plan, the Skills Union, and the upcoming roadmap for digital education and skills.

https://2eu.brussels/en/news-articles/european-commission-and-oecd-propose-that-students-learn-not-just-to-use-artificial-intelligence-but-to-understand-it-critically