The recent developments in AI-driven music generation have shown a significant shift towards integrating more complex musical elements and broadening the scope of cultural representation. Researchers are increasingly focusing on models that can not only generate music based on textual descriptions but also incorporate advanced musical features such as chord progressions, which enhance the coherence and richness of the generated compositions. Additionally, there is a growing emphasis on making these generative models more accessible through user-friendly interfaces, allowing for interactive and controllable music creation. However, a critical issue that has emerged is the underrepresentation of global musical diversity in AI datasets and research. Studies indicate a stark imbalance favoring music from the Global North, with genres from the Global South being significantly underrepresented. This lack of diversity poses a threat to the preservation and innovation of global musical traditions. Moving forward, there is a pressing need for more inclusive data collection and research practices to ensure that AI-driven music generation reflects and respects the rich tapestry of global musical heritage.

Noteworthy papers include one that introduces a novel approach to musical creativity using 2D cellular automata based on music intervals, and another that extends music generation models to include chord progression features, significantly enhancing the quality of generated music.