AI-Based Music Performance Assessment in Professional Music Education: A Critical Analysis from an Assessment Literacy Perspective

ABSTRACT

This study critically examines the adequacy of AI-based music performance assessment tools from an assessment literacy perspective and proposes a framework for their use in professional music education. Using Critical Interpretive Synthesis, 52 sources were analyzed through six dimensions of the Assessment Literacy Audit Protocol (ALAP): construct validity, reliability, fairness, formative utility, interpretive scope, and transparency. The findings indicate that AI tools are effective in consistently assessing technical elements such as pitch and rhythm and providing formative feedback for beginning learners, but remain limited in evaluating expressivity, interpretive individuality, and stylistic appropriateness. The study suggests integrating the bounded capabilities of AI assessment with human expert judgment.