As generative AI becomes increasingly present in artistic and musical creation, questions remain about how algorithmic compositions compare to human-written work at a structural and theoretical level. This project will investigate differences in harmonic motion, melodic contour, rhythmic density and complexity, large-scale formal organization, and approaches to orchestration and timbre across matched sets of AI-generated and human-composed pieces. Primary analysis may be conducted using production and notation tools available through my external advisor’s professional recording studio. Using excerpts composed under shared parameters, such as instrumentation, tempo, duration, and form, I will perform score-based and audio-feature analyses, and may supplement with computational metrics where appropriate. If time permits, I will extend this work through a small listening study in which participants will evaluate anonymized clips on criteria including emotional impact, originality, and perceived authorship, alongside brief memory-based measures.