Educational AI, Public Pedagogy, and the 2026 European Guidelines: The Strummin’ Gym Approach to Music Education

Abstract

The 2026 European Guidelines on the ethical use of artificial intelligence in teaching and learning provide a vital framework for navigating the nexus of cognition, learning, and technological mediation. This paper argues that the educational legitimacy of AI depends not on technical efficiency, but on its capacity to enrich the "cognitive ecology" of the classroom by supporting pedagogical judgment. By examining the structure of the Guidelines, the study shows how they contextualize the AI Act within the institutional ecology of schooling, advocating for domain-sensitive design grounded in embodiment and teacher mediation. Music education serves as a primary illustration, revealing non-binary and interpretative dimensions of learning – such as improvisation and stylistic accompaniment – that many current self-directed AI applications fail to address. In this context, the Strummin’ Gym framework is presented as a case study for integrating teacher-led mediation into AI-powered tools. By keeping the educator at the center of both the neural network's training and the pedagogical relationship, the framework moves beyond rigid binary assessments toward stylistically plausible support. This approach demonstrates how AI-assisted platforms can transform musical practice into a guided, reflective, and progressively autonomous experience, reclaiming the teacher’s role as the indispensable mediator of music as a living, relational art.