This quasi-experimental study aims to investigate the effectiveness of artificial intelligence (AIVA app) within an interactive music-learning environment. The ANCOVA was used to compare the performance of students using Artificial Intelligence Virtual Artist (AIVA) and those studying with traditional lecture-based instructions. Students in the AIVA group quickly mastered the new material and reported interest in the theoretical part of the course. The lecture group showed the opposite trend. Future research can involve a larger sample of 3-5 universities and compare student performance across Europe, Asia and the Americas. It is also worth considering the main shortcomings of the existing online artificial intelligence systems for music generation. The results of the study suggest that lecture-based instructions are less effective than artificial intelligence in teaching music. However, it seems important to have direct interactions during class; otherwise, students will lose interest, pay less attention and report fatigue.