Teachers and managers in vocational education and training (VET) are constantly asking the same question about artificial intelligence: What does it actually look like in practice? Policy documents and ethical guidelines provide necessary frameworks, but educators want concrete examples of how their peers are navigating this new terrain. A recent mapping commissioned by the European Commission's Directorate-General for Education, Youth, Sport and Culture offers exactly this, providing a much-needed spotlight on EU case studies and insights into the practical application of AI in education.

This document is particularly valuable because it cuts through the typical AI hyperbole, focusing instead on the reality of implementation. The authors began with a multilingual desk review that resulted in 110 verified practices, supplemented by a pan-European survey of educators, school leaders, and policymakers that brought in another 99 examples 1. From this combined pool of 209 practices, the researchers selected 14 cases for in-depth investigation. They chose cases based on their pedagogical core, their human-centered dimension, and their implementation reality. This, they say, ensures the featured examples are not just theoretical concepts, but actual tools being used, tested, and sometimes struggled with, in real classrooms.

A Diverse Range of Practices

The range of practices covered in the selected case studies is broad, spanning early childhood education to adult learning and higher education. This diversity demonstrates that AI's potential is not limited to advanced technical subjects or older students.

For instance, the mapping highlights Lalilo in France, an adaptive platform used in early childhood and primary education to build foundational reading skills. It uses speech recognition to offer personalized phonics and comprehension activities, allowing students to work independently and at their own pace. At the other end of the spectrum, the CurreChat project at the University of Helsinki provides a secure, generative AI tool for higher education students and staff 1. It allows for creative exploration and safe practice environments, such as simulating patient consultations for medical students.

In the VET and adult education spheres, the examples are highly pragmatic. The @2SOSUAI project in Denmark uses a generative AI chatbot named Amina to support female students with Danish as a second language in social and healthcare education. Amina acts as a round-the-clock learning partner, helping students reflect on placement scenarios and understand technical terms, directly addressing the language and cultural barriers that often hinder course completion. Similarly, the ROLEPL-AI project, also in Denmark, uses AI-supported role-play simulations to help adult learners practice workplace soft skills, such as communication and collaboration.

Common Threads and Notable Features

Despite the diverse contexts, several common threads emerge across these case studies. The most notable is the central role of the teacher. The survey data underscores this, revealing that teacher training and support are the most critical success factors for the successful application of AI. The case studies consistently show that AI tools are most effective when educators are involved in their design and implementation, rather than having technology imposed upon them from above.

It is notable that most of the projects started in attempts to solve existing problems, rather than dreaming of the purposed benefits of technology.

For example, the NOLAI co-creation laboratory in the Netherlands starts with concrete classroom challenges identified by teachers themselves, ensuring that the resulting AI solutions are genuinely useful in day-to-day teaching. In the ROLEPL-AI project, teachers had to build deep skills in prompting and content creation to design the simulations, highlighting that using these tools is not a simple time-saver but requires significant preparation and sustained effort.

Another striking commonality is the focus on critical AI literacy. Many of these practices do not just use AI to teach a subject; they teach about AI through its use. The DigiHavel tool in the Czech Republic, an AI chatbot used for civic education, explicitly teaches students that AI can produce inaccurate information, requiring them to think critically and verify answers. In the AI Twins project in Lithuania, where lecturers created digital twins of themselves based on their course materials, the final exam requires students to interrogate the Twin until it hallucinates and then critique the output 1. This approach turns AI reliability - or the lack of it - into a direct learning objective.

How These Insights Help VET Practitioners

For VET practitioners reading this mapping, the value lies in the realistic portrayal of both the benefits and the hurdles of AI integration. The document does not shy away from the challenges. It candidly discusses issues such as technical slowdowns, data protection complexities, the significant preparation time required for innovative AI teaching, and the risk of over-reliance by students.

However, it also provides actionable insights. The Italian pilot of AI-powered virtual assistance in vocational schools demonstrates the importance of comprehensive training and building AI tools directly into daily practice. The Danish @2SOSUAI project illustrates how AI can be targeted to support inclusion and address specific learner vulnerabilitie

Ultimately, these case studies offer VET educators a practical vocabulary for discussing AI. They move the conversation away from abstract fears or exaggerated promises and toward concrete pedagogical questions: How can we involve teachers in the design process? How do we ensure data privacy? And most importantly, how can we use these tools to foster critical thinking and support the specific, practical needs of our learners? By studying these real-world examples, VET institutions can better navigate the complexities of AI adoption and focus on what truly matters: enhancing the quality and equity of vocational education.

References

Crêteur, S., Dunajeva, J., Lazaro Soler, M., & Siarova, H. (2025). *GOOD PRACTICE IN AI FOR EDUCATION: Spotlight on EU Case Studies and Insights*. Mapping commissioned by the European Commission, Directorate-General for Education, Youth, Sport and Culture.