<p>This study re-examines the <i>Arca Musarithmica</i>, a musical algorithmic device developed by Athanasius Kircher in the mid-seventeenth century, through the perspectives of contemporary music theory and artificial intelligence (AI)-based music generation. The primary aim is to elucidate the theoretical and practical significance of Kircher’s algorithmic approach in modern music theory and AI music generation models. To this end, this study addresses three specific research questions: (1) How can Kircher’s <i>Arca Musarithmica</i> be interpreted as a historical precedent for explainable AI systems in music? (2) How does the explicit, pattern-based combinatorial algorithm of Kircher offer conceptual insights for addressing the “black-box” problem in contemporary AI music generation? (3) What insights can Kircher’s framework offer for shaping future approaches to music composition and education, regarding transparency and interpretability? To address these questions, this study analyzes Kircher’s original documentation in <i>Musurgia Universalis</i> (1650) and evaluates its harmonic structure using functional harmony. The findings suggest that Kircher’s device parallels the principles of voice leading and harmonic progression later codified in functional harmony, while still grounded in seventeenth-century contrapuntal and rhetorical traditions. Furthermore, this study establishes that Kircher’s explicit, pattern-based combinatorial algorithm offers a promising foundation for improving interpretability in modern AI-based music models. Ultimately, this study provides a theoretical framework for developing human-centered, interpretable AI music models by integrating historical musical knowledge with contemporary technology, greatly benefiting music composition and educational practices. This interdisciplinary exploration innovatively bridges historical algorithmic composition techniques with contemporary AI methodologies, opening new pathways for interpretable and human-centered music generation.</p>