[White paper] An Overview of Music AI and its Potential (2025 edition)

Qosmo Inc., a company focused on utilizing AI in creative fields, has released a free white paper, "An Overview of Music AI and its Potential (2025 Edition)", which outlines the overview of AI technology in the music domain—an area that has seen particularly remarkable development in recent years—along with its applications in industry. This report is a sequel to the well-recieved “An Overview of AI Music Generation and its Potential (2022 Edition)”, published in March 2022.

This update focuses on music generation AI (AI technology that generates music predominantly from text prompts), which has seen especially dramatic technological advancement in recent years, while also covering a wider range of topics including generative AI technologies that assist artists in music production and AI for music analysis. At the same time, the report also addresses ethical aspects of AI technology, such as copyright infringement issues with training data.

The first half (Chapters 1-3) summarizes specific application examples and use cases, making it enjoyable content even for those without technical knowledge. For those who want to learn more about technological trends, please also refer to Chapter 4.

Download

(日本語版はこちら/Japanese version is also available.)

Contents

Introduction About Qosmo Ch 1 — About Generative AI in Music Ch 2 — AI tools Supporting Music Production Ch 3 — AI that Understands and Analyses Music Ch 4 — Cutting-Edge Music AI Conclusion

Authors

Naotake Masuda Graduated from the University of Tokyo Graduate School of Engi- neering, Department of Electrical Engineering, Doctor of Engineering in 2023. In 2019, worked as an intern at IRCAM (Institute National de Recherche Acoustique et de Music) in France. Mainly researched deep learning models and genetic algorithms that support synthesizer operation and synthesize new sounds. At Qosmo and Neutone, he is involved in R&D and consulting with client companies as an AI researcher and engineer, and is also involved in the development of new models for Neutone.

Tom Baker With an academic background in mathematics and physics, Tom joined Qosmo in spring 2025 to explore his artistic interests in aug- mented music interfaces and sound design. He is focused on lever- aging machine learning as another tool to expand the sonic palette available to musicians and create deeply playable electronic in- struments, both software and hardware. Tom is currently pursuing a PhD at the University of Manchester and has previously worked on projects at Sony CSL Paris.

Nao Tokui Artist, researcher, and CEO of Qosmo Inc., explores the enhancement of human creativity through AI from both research and creative production perspectives. Leads Qosmo, a collective composed of artists, designers, AI researchers/engineers, engaging in creative work and technology development. Additionally, through Neutone, established in July 2023, he is developing new “instruments” using AI. His major publication includes “AI for Creation — The Endless Story of Machines and Creativity” (recipient of the Okawa Publication Award). Completed doctoral program in Electronic Engineering at the University of Tokyo, Graduate School of Engineering. Ph.D. of Engineering.

Design Liu Jinhe