Interactive learning material

Active Learning and Developing Transversal Skills with Artificial Intelligence

How does AI-supported active learning improve critical thinking, collaboration, communication, creativity and learner autonomy?

Author: Mária HartyányiiTStudy Hungary Training and Research Centre
Reviewer: Annamária KacsuriTStudy Hungary Training and Research Centre

AI connects inquiry, communication, collaboration, creativity and independent learning. — The student at the centre: dialogue with peers and with AI

Learning Objective

The essence of active learning is that students are not recipients but shapers of the learning process: they ask questions, investigate, experiment, create, argue and reflect. The emergence of generative artificial intelligence (AI) has established a new situation. Experience so far and the first research findings show that AI can be particularly effective in supporting active learning – provided that the teacher integrates it into the teaching process consciously and subordinated to the learning objectives.

In this learning unit, we present how artificial intelligence (AI) can be combined with three proven active learning methods: inquiry-based and research-based learning, the flipped classroom, and project-based learning (PBL). After working through the material, teachers will be able to:

explain why these methods can be more effective with AI than before;
design AI-supported pre-class activities, research questions, project scenarios and reflection processes;
use AI in a way that improves students’ critical thinking, collaboration and autonomy.

Chapter 1 · Skills

Transversal skills – Why is now the time?

Transversal skills are abilities that are independent of any specific field and are needed in every profession and life situation. The best known of these are the “4Cs” of 21st-century learning: critical thinking, communication, collaboration and creativity. Closely related to these is learner autonomy, i.e. the ability of students to plan, monitor and evaluate their own learning. The European Union’s Recommendation on Key Competences (2018) and the LifeComp framework describe the same thing as the competence of “learning to learn”, while DigCompEdu, which describes the digital competences of educators, describes it as supporting collaborative and self-regulated learning and as empowering learners.

Critical thinking

We ask good questions, seek evidence, check answers and build arguments.

Communication

We express ourselves clearly, listen to each other and give understandable answers in professional situations.

Collaboration

We decide together, share responsibility and bring different perspectives together.

Creativity

We develop alternative ideas, compare them and create new solutions.

Learner autonomy

Students plan, monitor and evaluate their own learning. This ability is related to the 4Cs and requires independence and self-regulation.

Chapter 1 · The roles of AI

Why does it work better with AI than before?

If a fact, a summary or a piece of program code can be produced in seconds, the value of knowledge shifts: what matters is whether the student can ask a proper question, check the answer, defend position, and create something new together with others. Active learning situations make them practise exactly these skills – and AI helps effectively in planning, designing and managing these situations:

Socratic partner

as a Socratic partner, it asks follow-up questions and raises counterarguments – this develops critical thinking;

Simulator

as a simulator, it plays the role of the client, the guest or the expert – this develops communication;

Idea generator

as an idea generator, it offers alternatives that the group has to decide on together – this develops collaboration and creativity;

Tutor and reflection partner

as a tutor and reflection partner, it adapts to the student’s own pace and helps them think through the learning process – this develops learner autonomy.

Though active learning has long been known, yet it appears in everyday teaching practice less often than we would like: it requires a lot of preparation, makes differentiation difficult, and in large groups there is little time for individual feedback. According to international experience, this is exactly where AI helps. In England, teachers using ChatGPT spent nearly a third less time on lesson planning (EEF, 2024). At the same time, the OECD’s 2026 report warns that while general-purpose chatbots may improve immediate performance, but not necessarily learning, if students let thinking to them as well.

Pedagogical design makes the difference

Pedagogical design determines whether AI strengthens active learning by surging students to think, or works instead of them.

Chapter 2 · Inquiry

Inquiry-based and research-based learning with AI

Depending on the role the teacher takes in preparing and conducting the research, and on how independently the students work, the literature uses two terms. In inquiry-based learning (Inquiry Based Learning, IBL), the questions are typically asked by the teacher, who also guides the discovery, whereas in research-based learning (Research Based Learning, RBL), students are given more independence and their work follows the steps of scientific research more closely. The phases of the work are the same in both approaches; the difference lies in the role of the teacher and the degree of student independence. AI can support both students and teachers in every phase – but it can never take over the work of curiosity, questioning and verification.

The student’s question is followed by dialogue with AI, investigation and an evidence-based conclusion. — Ask, investigate, prove – when AI questions back!

Inquiry-based learning (IBL)

The teacher typically asks the questions and guides the inquiry.

Research-based learning (RBL)

Students have greater independence; their work follows the steps of scientific research more closely.

Chapter 2 · Research phases

How can AI help in each research phase?

Orientation
with the help of AI, the teacher can create a realistic case description, a short scenario or provocative questions that arouse students’ curiosity. The solution is deliberately left out here.
Questions and hypotheses
students first formulate their own questions and hypotheses, and then ask AI to question them back, look for counterarguments and help them refine the question.
Investigation
AI can review the design of the experiment or survey (What are we measuring? What is the benchmark?), help draft a questionnaire, organise data into tables and create charts. The measurement, the fieldwork and the experiment, however, are carried out by the students.
Conclusion
acting as the “devil’s advocate”, AI can offer alternative explanations, and students have to decide, using arguments, which interpretation their data support. Checking sources is also part of the task at this stage.
Sharing
AI can help structure the presentation, make the professional text easier to understand or prepare a summary in a foreign language, but presenting and answering questions remain the students’ tasks.

The question belongs to the student

The most important rule: questioning remains with the student. If AI formulates the assumption (the hypothesis), the research loses its pedagogical purpose. If, however, students “test” their own hypotheses with AI and have to defend or modify them, critical thinking and reasoning will improve.

Chapter 2 · An example from baker training

Why won’t the dough rise?

A group of 10th-grade baker students is investigating the proofing process. The teacher does not start with an explanation but with a problem. The students formulate hypotheses (temperature, yeast, salt), carry out an experiment in the training workshop, and, based on their results, write a professional guide for a bakery.

Teacher prompt

write a short, realistic case description for a group of 10th-grade baker students about a small bakery where, over the past week, the bread dough has not been rising properly in the mornings. Include a few telling details (for example cold weather, a new yeast supplier, a new staff member), but do not give the solution! Add five open-ended questions that encourage the students to formulate hypotheses.

Student prompt

our hypothesis is that the dough does not rise because the bakery is too cold. Don’t tell us whether we are right! Ask three questions that help us refine our hypothesis, and suggest a simple experiment that we can carry out in the training workshop within two hours. What do we need to measure, and what should the control be?

Chapter 3 · Flipped classroom

Flipped classroom with AI

The flipped (or inverted) classroom (flipped classroom) is a student-centred method. The “flip” means that working through the learning material does not begin with the teacher’s explanation, but with students studying the digital content provided by the teacher (video, text, interactive task) before class, at their own pace. This leaves more time in class for joint work, problem-solving, discussion and practical application.

According to teachers’ experience, the method has three common obstacles: preparing the pre-class material takes a lot of time, not every student works through it, and the teacher cannot see exactly who has understood what before the lesson. AI offers a solution to all three problems.

At home, students prepare independently using digital content and AI; in class, the group works together. — Individually at home, together in class

Making use of the extra time in class

Joint work, problem-solving, discussion and practical application: meanwhile, the teacher observes, asks questions and gives feedback.

Chapter 3 · The learning process

Before, during and after class

Before class – preparation

With the help of AI, the teacher can quickly draft the pre-class material: a short explanatory text, a video or interactive learning material, a glossary and/or check questions that can be embedded in the material. The same content can be produced at several levels, which supports differentiation. However, the professional accuracy of the content must always be checked by the teacher.

Students can use AI as a “patient tutor”: if they get stuck, they can ask it questions, but in such a way that it does not give the ready-made answer but helps them with guiding questions. Part of the pre-class task may be for each student to bring two or three questions of their own to class. If the teacher summarises the results of the quiz and the students’ questions with AI – without any personal data – they can already see the typical mistakes before the lesson and can build the lesson on them.

In class – thinking together

Since the basic knowledge has already been acquired before class, in class the students interpret and apply it in groups. Here, AI appears primarily as a simulator and debate partner: it plays the role of a client or guest, formulates opposing views, or gives the group further examples. Meanwhile, the teacher moves between the groups, observes, asks questions and gives immediate feedback – no tool can replace the presence of the teacher.

After class – deepening learning

After class, students receive a creative task (for example, they prepare a plan, a guide or a short presentation), on which AI can give formative feedback based on predefined assessment criteria. The process ends with a short reflection: what did I learn, what was difficult, how did AI help – or where did it mislead me?

Chapter 3 · An example from hospitality

Allergens in the dining area

A group of 11th-grade cook and waiter students is learning about food allergens and how to inform guests. EU regulations define 14 major allergens about which guests must be informed.

Pre-class learning

Before class: students watch a short video and complete the embedded quiz. Using AI, the teacher also attaches a glossary and a simplified version of the text to the video. Students can practise with AI acting as a tutor and bring two questions to class.

Teacher preparation

Preparation: from the quiz results, the teacher sees that most mistakes relate to “hidden” allergens (sauces, dressings, coatings), so builds case tasks around these.

In-class application and role-play

In class: the groups receive a real menu and have to decide which dishes contain what allergens. This is followed by a role-play: AI acts as a guest with a peanut allergy, and the students answer as waiters. Another group acts as observers and evaluates the communication: was the answer accurate, polite and safe?

After-class creation and reflection

After class: the groups prepare the allergen labelling of the menu and a short information sheet for guests, and then write an individual reflection.

Chapter 4 · Project-based learning

Project-based learning (PBL) with AI

Project-based learning is a teaching and learning method in which the project provides the framework for learning, but every activity is linked to the learning objectives set out in the curriculum. Students work on a real problem related to their own experiences and future profession, plan and decide together, and at the end of the work present a tangible outcome – a product, service, plan or presentation – preferably also to an audience outside the class. PBL is particularly useful in vocational education and training because it brings school closer to the world of work and enables students from several vocations to work together on a topic. The project consists of well-defined phases, and AI can support both teachers and students in each of them.

The project team follows milestones, creates together and works towards its goal with AI support. — Creating together: the team decides, AI helps

Driving question and scenario

a real, open problem linked to the learning objectives.

Planning

roles, tasks, milestones and timeline.

Research and creation

information gathering, experimentation, prototyping, data analysis.

Feedback and revision

peer and self-assessment, teacher and AI feedback, further development of the work.

Presentation

public presentation, preferably also to an audience outside the school.

Reflection

what we learned, how we worked together, how we used AI.

Chapter 4 · The teacher’s perspective

The teacher as a project designer

The effectiveness of project-based learning is determined primarily by pedagogical design. In this work, AI can help as a pedagogical assistant to:

suggest driving questions and realistic project scenarios aligned with the learning objectives;
break the project down into phases, milestones and tasks, and draw up a timeline;
develop differentiated roles and tasks for students with different interests and abilities;
create assessment criteria and assessment tables (rubrics) for peer and self-assessment;
summarise the groups’ interim reports and reflections, and indicate where teacher’s intervention is needed.

Teacher prompt · a three-week project

I teach at a vocational school. Design a three-week interdisciplinary project for mixed groups of gardener, electrician and IT students on modernising the irrigation of the school garden. Provide a driving question, the phases and milestones of the project, the roles within the group, and a four-level assessment table (rubric) that, in addition to the professional outcome, also assesses collaboration, communication and creativity. The finished plan is only a starting point, which we will finalise together with the students.

Planning together with students

AI’s suggestion is a starting point. The teacher reviews the driving question, roles, milestones and assessment criteria, then finalises them with the students.

Chapter 4 · An interdisciplinary project

The team decides, AI helps

The students’ perspective

Working on a project, students learn to interpret the problem together, plan the work, divide the tasks and share responsibility. AI can help them as a conversation partner and a working tool:

in brainstorming and comparing different solution options;
in searching for, comparing and verifying information;
in analysing measurement data and creating tables and charts;
in writing and debugging program code, with AI also explaining how the code works;
in structuring the presentation, creating visual elements and writing a summary in a foreign language.

Example: irrigating the school garden

Example (interdisciplinary project): mixed teams of gardener, electrician and IT students investigate how the water consumption of the school garden can be reduced by a third. They design and build irrigation controlled by soil moisture sensors, analyse the measured data, and present the results to the school management and a local horticultural entrepreneur.

Student prompt · deciding together

These are our three ideas for reducing the garden’s water consumption: … Don’t choose for us! Ask two critical questions about each of them, and help us put together a set of criteria (cost, feasibility, impact) on the basis of which we can decide as a team.

Chapter 5 · Feedback and autonomy

Feedback and reflection with AI

A common element of all three methods is continuous feedback and reflection. Based on predefined assessment criteria that are also known to the students, AI can give formative feedback during the work, so students can make improvements several times.

Shared assessment criteria

Using AI, the teacher can draft the assessment table (rubric), which they then discuss and finalise together with the students.

Improvement before submission

Students can improve the draft of their work with AI feedback based on the assessment criteria before submitting it.

Constructive peer assessment

Before peer assessment, AI can help formulate constructive, specific feedback (“What went well? What could be improved? What do you suggest?”).

Recognising difficulties

AI can summarise the groups’ reflections and alert the teacher to recurring difficulties.

“How did I use AI?” log

In a short “How did I use AI?” log, students describe what they used AI for, which suggestions they accepted, which they rejected, and why. Reflection contributes to the development of their digital competences and at the same time strengthens their autonomy. Responsibility for assessment and grading remains with the teacher.

Student prompt · reflection

We have just finished the project. Ask me five questions, one at a time, about what I learned, what was the most difficult, how we worked together, and what I used AI for. Don’t evaluate my answers; just ask follow-up questions if I am too general. At the end, summarise in three sentences what I should do differently next time.

Chapter 6 · Takeaways and practice

Takeaways for teachers

Applying artificial intelligence in education does not require developing a new, “AI-based teaching method”. Introducing AI can make a reality of what progressive pedagogy has been advocating for more than a hundred years: that students ask questions, investigate, create, work together and take responsibility for their own learning.

When digital tools were introduced, many hoped that technology would revolutionise teaching “by its mere presence”. It didn’t happen so. The interactive whiteboard, the digital textbook or the projector made the traditional, frontal lesson more colourful in most cases, but did not change the logic of learning. One reason for this is that the real obstacle to introducing active learning methods was not a lack of technology, but a shortage of teachers’ time and energy. Preparing for a flipped lesson, a project task or differentiated group work requires much more work than traditional explanation. Early digital tools did not reduce this burden; on the contrary, they often added new tasks to it.

This is exactly where AI makes a difference. In a few minutes, the teacher can create tasks of different difficulty levels, guiding questions, debate-starting statements or case studies, which significantly reduces preparation time. In addition, AI can provide something that a teacher in a class of 30 cannot: immediate, personalised feedback and support for every student. Students can ask questions, try things out and make mistakes at their own pace, while the teacher freed of these can support discussion, guide thinking and facilitate joint work in class.

Of course, none of this happens automatically. AI can just as easily serve old, passive learning if students only ask it for ready-made answers. The real value of this opportunity therefore lies not in the tool itself, but in the pedagogical approach with which we use it.

The real value of this opportunity lies in the pedagogical approach with which we use AI.

Exercise 1 · From a frontal lesson to an active lesson

30 minutes · individual preparation

Choose one of your own lessons and use AI to redesign it for inquiry-based learning or the flipped classroom. Specify what students, the teacher and AI do; check professional accuracy and plan a reflection activity.

Exercise 2 · “Don’t tell us the answer!”

45 minutes · groups of 3–4

The group first formulates a hypothesis about a professional problem without AI. After AI’s questions and counterarguments, students refine or defend it together, plan an investigation, present their reasoning and reflect.

The detailed steps and accompanying prompts are in the practical exercise available to download from the final page.

Chapter 7 · Knowledge check

Check your knowledge!

Choose one answer for each question, then check your answers. Restart to clear your answers and feedback.

1What does learner autonomy mean?The student plans, monitors and evaluates their own learning.The student leaves every decision to AI.The student learns only from teacher explanations.

2Who owns the research question and hypothesis?AI, because it formulates them faster.The students; AI can ask follow-up questions and raise counterarguments.Always an external expert.

3What is the essence of the flipped classroom?The teacher reads out AI’s answers in class.All joint work must be done at home.Students work through the basic knowledge before class and apply it in class.

4How does the project team use AI’s ideas?It asks for criteria and critical questions, then decides together.It automatically chooses the first idea.It asks AI to take over responsibility.

5Who is responsible for assessment and grading?AI alone.The teacher.The manufacturer of the device used to access the chatbot.

6What should the “How did I use AI?” log include?Only the name of the chatbot.The entire conversation without any explanation.What I used AI for, what I accepted or rejected, and why.

Chapter 8 · Further reading

Credits and copyright

Author: Mária HartyányiiTStudy Hungary Training and Research Centre
Reviewer: Annamária KacsuriTStudy Hungary Training and Research Centre
Project identifier: VITA · M3-T2-MLC1

Attachments

Learning material source (PDF)M3-T2-MLC1-TEXT-EN.pdf Practical exercise (PDF)M3-T2-MLC1-EXC-EN.pdf

Editable Word version of the practical exercise

This work may be reused and adapted under the terms of the Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0).

Images without a source attribution were generated by Claude.

Funded by the European Union. Views and opinions expressed are those of the author(s) only and do not necessarily reflect those of the European Union or the National Agency.