UNIZG FOI develops a comprehensive AI framework for European schools through the highly competitive Horizon project TRIAD

How can artificial intelligence (AI) be used in schools to help students learn more effectively, develop critical thinking and solve problems independently? The University of Zagreb Faculty of Organization and Informatics (FOI) is helping to answer this question by working with eleven partners from seven European countries to develop a comprehensive framework for the use of AI in education. The TRIAD project brings together educational sciences, artificial intelligence and neuroscience to investigate how learning unfolds with AI support and translate the findings into practical guidelines and teaching scenarios for schools across Europe.

This three-year research project, selected through a competitive funding process under Horizon Europe,  has received an EU contribution of almost €3.4 million. The consortium brings together 11 organisations and is coordinated by RPTU University Kaiserslautern-Landau in Germany. This collaboration combines research expertise with the experience of educational institutions and technology partners to develop solutions applicable across different European school systems.

Full Professor Nina Begičević Ređep, PhD, who leads the FOI project team, explains that the project addresses a challenge increasingly faced by teachers and parents: students have access to tools that can write a text, provide a solution to a task or explain a complex concept in seconds. However, a quick answer does not necessarily mean that students have acquired knowledge or developed cognitive skills such as critical thinking and problem-solving. The researchers will therefore examine how and under what conditions AI can support the development of cognitive skills, and when it encourages excessive reliance on ready-made solutions and AI tools. They will also explore how the use of technology and artificial intelligence affects young people’s emotional and social well-being. She emphasises that the framework will cover all levels of AI use in schools, from conceptual and strategic guidelines to practical teaching scenarios. She also highlights the importance of FOI’s participation in an international consortium of researchers and institutions with internationally recognised achievements in artificial intelligence and education.

A particular strength of the project is its use of neuroscientific methods, including the measurement of brain activity. By combining these methods with classroom research, the researchers will seek to better understand when AI tools support learning processes and when student engagement does not translate into the development of cognitive skills. This will provide decision-makers and those developing strategies with a scientific basis for determining how and in which situations artificial intelligence should be used.

The research focuses on students aged 12 to 18, with pilot studies involving approximately one thousand students in Germany, Lithuania, Finland, Greece, Italy and Croatia. The framework will be tested in real school settings, linking literacy, mathematical and scientific skills with critical thinking, collaboration, and digital and AI literacy. Monitoring student well-being and ensuring that the approaches developed are accessible to students with diverse needs and capabilities will also be central to the research.

Generative artificial intelligence has already become part of everyday life for many teenagers, yet schools are still exploring ways to integrate it meaningfully into education. The lack of such systematic frameworks raises key questions: how can artificial intelligence support learning? Do AI tools hinder students’ development of critical thinking, or can we use them to strengthen it and learn how they can support decision-making and problem-solving? How does technology use affect adolescents’ emotional and social well-being?

To answer these questions and provide systematic support to educational institutions, researchers will spend the three-year project period developing an integrated competence framework that connects foundational subject skills (literacy, mathematical and scientific skills), transversal competences (critical thinking, collaboration and learning skills), AI competences and indicators of student well-being. This framework will be tested in real school settings through pilot studies involving approximately one thousand students from schools in Germany, Lithuania, Finland, Greece, Croatia and Italy.

FOI participates in all project work packages, with a particular contribution to developing policies and guidelines to support decision-making on the use of AI in European schools. It will also help translate research findings into educational practice. Its team will contribute to the development of practical resources and tools, teacher professional development, the development of micro-credentials, and recommendations for the sustained and wider application of the project’s results. FOI has been allocated €212,500 in EU funding for its participation.

The consortium’s kick-off meeting took place on 21–22 September 2026 in Kaiserslautern, where the partners agreed on the research approach and initial activities. The project, whose full title is TRIAD – Balancing Foundational and AI Competences for Inclusive, Well-being-Oriented Education, began on 1 September 2026 and will run until 31 August 2029.

Expected outputs include a comprehensive AI framework for European schools, policies, strategies and strategic recommendations, teaching scenarios, and digital toolkits for students and teachers. These outputs aim to help schools use artificial intelligence thoughtfully while developing students’ cognitive skills, including decision-making, problem-solving and critical thinking, and ensuring inclusivity and the well-being of all stakeholders in an era of disruptive technology.

Alongside the FOI project team leader, the team includes Full Professor Neven Vrček, PhD; Associate Professor Nikola Kadoić, PhD; Associate Professor Goran Hajdin, PhD; Assistant Professor Bojan Žugec, PhD; Petra Vondra, MSc in Informatics and Josipa Bađari, prof.

Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Executive Agency (REA). Neither the European Union nor the granting authority can be held responsible for them.