The European Commission’s new Education and Skills Package puts AI literacy firmly on Europe’s educational agenda. The challenge now is turning an ambitious framework into knowledge that young people can actually use.
On 5 November 2026, the European Commission presented its new Education and Skills Package, addressing four connected priorities: strengthening basic skills, improving the use of digital technologies and AI in education, supporting the teaching profession, and modernising vocational education and training.
One proposal deserves particular attention: an AI Literacy Certificate for 15-year-old learners, with a pilot planned for the 2027/2028 school year.
The proposal builds on the EU–OECD AI Literacy Framework introduced earlier this year. It is an encouraging development because it recognises that artificial intelligence has become an educational subject in its own right. Young people need to understand technologies that increasingly influence how they learn, communicate, access information and participate in society.
But the certificate also introduces an important implementation question. How do we ensure that the knowledge Europe wants to recognise is knowledge our education systems are genuinely prepared to teach?
At REDefine, we believe this is where some of the most consequential work will take place.
What does the proposed EU AI Literacy Certificate actually involve?
The European Commission plans to pilot an AI Literacy Certificate for 15-year-olds during the 2027/2028 school year. This is a proposed initiative, not an existing mandatory Europe-wide qualification.
Its educational foundation is the EU–OECD AI Literacy Framework for Primary and Secondary Education, which identifies 19 competences across four dimensions: engaging with AI, creating with AI, managing AI and shaping AI.
Importantly, the framework is not simply a guide to prompting chatbots. It addresses technical understanding, critical evaluation, responsible use and the wider ethical and social implications of artificial intelligence.
That distinction matters. A young person can become highly proficient with a particular AI application without understanding the systems, incentives or limitations behind it. Meaningful AI literacy must reach further.
The framework recognises this. The next challenge is translating those competences into ordinary educational experiences, across different schools, communities and levels of preparation.
Familiarity with technology is not the same as understanding it
Through REDefine’s educational activities, we work with young people from different countries and socioeconomic backgrounds. We use introductory activities to establish shared foundations before progressing to more complex learning about technology, digital safety and responsible AI.
Something repeatedly surprises us: many participants are confident users of digital platforms, and increasingly AI applications, yet the majority in a number of our groups are encountering at least some foundational concepts for the first time.
These include the commercial value of behavioural data, how algorithmic classification works, why automated decisions can reproduce discrimination, and the limitations of AI-generated information.
These are observations from our educational practice, not representative survey findings about European young people. Nevertheless, they reveal a practical problem that educators and curriculum designers should take seriously.
We have spent years watching younger generations become increasingly comfortable with digital interfaces. That comfort can create the impression that they already possess the underlying knowledge.
But knowing how to navigate a platform tells us relatively little about whether someone understands the decisions the platform makes, the information it collects or the consequences of those decisions.
This is one reason the proposed certificate deserves a thoughtful implementation strategy.
Three layers of digital understanding that education needs to connect
In our work on digital learning, we find it useful to distinguish between three connected forms of knowledge.
Interface literacy concerns what a person can do with a technology. Can a learner use an AI assistant, navigate a digital platform, organise information or work with a particular application? These are useful capabilities, but they are only the visible layer.
System literacy concerns what the technology is doing beneath that interface. What information does it process? How are outputs generated or ranked? What can the system infer about a person? What limitations, commercial incentives or design choices influence its behaviour?
Agency literacy concerns what a person can do with that understanding. Can they recognise when a system is unreliable, question a recommendation, protect their information, challenge a decision or decide that human judgement is necessary?
Consider a simple illustrative classroom activity. Students examine a digital platform that recommends different job opportunities to different users. Rather than asking only whether the recommendations are useful, the educator invites them to investigate what information might influence the results, how unequal visibility could affect opportunities and what someone could do if an automated decision appeared unfair.
The activity moves beyond operating a tool. It introduces the relationships between data, classification, opportunity and accountability.
This is not an alternative to the European AI Literacy Framework. It is one practical way to think about how its ambitions can connect to broader digital and civic education.
We explored this distinction further in our earlier article, The Code Beneath the Floorboards: The Open School.
Teacher preparation is part of the educational infrastructure
One of the encouraging aspects of the new Education and Skills Package is that the Commission has not treated AI literacy as an isolated student competence. The package also addresses teacher preparation, professional development and the wider conditions in which educators work.
The proposed next generation of Erasmus+ Teacher Academies, beginning in 2027, is among the measures intended to help teachers respond to emerging challenges, including AI.
This recognition is essential.
For many years, digital professional development has understandably concentrated on helping teachers use educational technologies effectively. Learning platforms, digital resources, collaborative tools and online teaching methods have all required considerable adaptation.
But understanding how to use a technology for teaching is different from being prepared to teach young people about the systems behind that technology.
A teacher might be highly capable of incorporating an AI-assisted activity into a lesson while having received little opportunity to explore behavioural profiling, algorithmic fairness, data economies or the relationship between automated decisions and fundamental rights.
These subjects also cross traditional disciplinary boundaries. They belong partly to computing, but also to citizenship education, economics, social sciences, media literacy and ethics.
That makes curriculum integration just as important as technical teacher training.
If educators are expected to develop these competences among young people, they need more than access to tools or occasional demonstrations. They need coherent learning materials, opportunities to deepen their own conceptual understanding, appropriate teaching time and methods for assessing whether students can apply what they learn.
What should schools and education providers prepare for?
The proposed certificate creates an opportunity to examine what meaningful AI literacy education looks like before assessment becomes the dominant conversation.
A useful starting point is to understand learners’ existing knowledge rather than assuming competence from frequency of technology use. A young person who uses AI every day may still need introductory learning about data protection, model limitations and automated decision-making.
The next step is to design coherent learning pathways. A single awareness workshop can introduce important questions, but deeper understanding requires opportunities to revisit ideas, examine cases, practise judgement and connect knowledge to different settings.
Teacher development must progress alongside curriculum design. Otherwise, ambitious learning outcomes risk becoming another responsibility assigned to educators without sufficient support to fulfil it.
Assessment matters, too. If AI literacy includes critical understanding and responsible decision-making, its evaluation should be able to recognise those capabilities rather than concentrating exclusively on producing successful outputs with AI tools.
Finally, implementation must pay attention to the differences between educational environments. Schools do not begin with identical resources, professional-development opportunities or institutional capacity. A European framework can establish common ambitions while still requiring substantial adaptation and support at local level.
None of this diminishes the value of the Commission’s proposal. It explains why the work of implementation deserves as much attention as the announcement.
From European policy to meaningful learning
There is a larger educational change behind this discussion.
Knowledge that might once have been regarded as specialist ICT education is becoming increasingly relevant to general education. Digital systems already shape employment, public services, access to information, economic participation and democratic life.
Understanding these systems is no longer a concern only for people who intend to study computing.
That is why REDefine welcomes the European Commission’s initiative. The proposed AI Literacy Certificate, supported by the EU–OECD framework, provides an important opportunity to strengthen how young people understand and navigate an AI-shaped society.
But the opportunity will be realised through education, not certification alone.
Read our full analysis: In Europe Wants to Certify AI Literacy. What Will It Take to Teach It?, we examine the implementation challenge in greater depth, drawing on REDefine’s educational experience and exploring the implications for teacher preparation, curricula and meaningful assessment.
The article also includes a free introductory learning resource developed through our CareLink work, offering educators and youth workers a practical starting point for conversations about digital safety, AI and responsible support.
Working on AI literacy, curriculum development or responsible digital education?
For schools, education providers, public institutions and European project partnerships, the challenge is increasingly how to translate complex knowledge into educational experiences that people can understand and use.
REDefine works at the intersection of learning design, civic education, responsible technology and digital learning environments. Our approach connects conceptual foundations with practical activities, accessible learning resources, appropriate technology and the institutional conditions needed to make educational programmes work.
If your organisation is developing an AI literacy programme, designing training for educators or preparing a European educational cooperation initiative, we welcome conversations about how meaningful implementation can be designed.
Contact REDefine at info@redefine.pt to discuss potential cooperation.
Official sources
European Commission — Education and Skills Package, November 2026
European Commission — Shaping Europe’s future through education and training, 5 November 2026
European Commission — EU–OECD AI Literacy Framework, June 2026
