Artificial intelligence (AI), due to its extensive capabilities for analyzing incoming information and providing personalized and detailed feedback, has significant potential for use in education. This study aims to investigate an application that analyzes the performance of musical exercises and provides instant AI-based feedback. The main task is to explore the impact of using this application on learning, as well as on the indicators of self-efficacy and academic locus of control. These indicators reflect an individual's beliefs about their ability to determine their own success in learning activities. According to the results, in the experimental group that studied using instant feedback provided by AI, the self-efficacy score on the Self-Efficacy for Musical Learning scale increased from 51.13 to 57.38. A dependent sample t-test confirmed the significance of differences at the level of p = 0.012. In turn, the control group studied according to the classical curriculum and showed no significant changes (53.46 in post-test vs. 52.78 in pre-pest, p = 0.756). Additionally, in the experimental group, the total score on the test of musical skills significantly increased from 19.89 to 34.08 out of 50, compared to an increase from 20.13 to 28.26 in the control group. A Student's t-test was used to compare the post-test results between the experimental and control groups. The comparison revealed significant differences in self-efficacy and musical skills in favor of the experimental group. At the same time, there was no significant effect of the studied educational intervention on academic locus of control measured using the Academic Locus of Control Scale. The results can be useful for curriculum developers and researchers studying the impact of AI technologies, particularly those that provide instant feedback for error correction. Further research may focus on exploring the longer-term effects of regular use of AI-based feedback in music training.