The rapid advancement of generative AI technologies is increasingly accompanied by their broader application across various sectors, particularly within creative industries. In this study, IBM Watson BEAT software, based on generative AI technology, was tested as an educational tool in the process of teaching music composition theory to flute students. The Revised Two Factor Study Process Questionnaire (R SPQ 2 F) was employed to investigate predominant learning approaches among study participants. Comparison of pre- and post-test results using the paired and independent samples t-statistics demonstrated significant improvements among students in the experimental group in the deep motive index (increase from 11.74 to 12.86, t = -2.013, p = 0.000) and the surface motive index (decrease from 12.87 to 11.66, t = 2.213, p = 0.002), which correspond to motivational foundations for choosing learning strategies, as well as in the complexity index of the musical compositions created (increase from 3.56 to 4.79, t = -2.146, p = 0.000). However, no significant impact on indices corresponding to practical strategies of learning behavior (deep strategy and surface strategy) was observed, indicating that the duration of the educational intervention may have been insufficient to induce lasting changes in study habits. The obtained results suggest a trend toward a positive influence of generative AI technologies on the formation of students' learning strategies. The results may prove beneficial to researchers investigating the integration of AI technologies into education and to those conducting studies aimed at modernizing and enhancing the effectiveness of current educational programs.