Sony Group has reportedly trawled out some new tech aimed at detecting original music played in AI-composed songs. The technology is intended to enable rights holders to create the copyrighted content included in AI-generated tracks and understand whether their content has been used without consent. With a number of lawsuits against AI music startups and increasing worries about the transparency of training data, the technology is being framed as a step toward addressing ethical and legal challenges posed by AI-generated music.
Sony’s AI Music Detection Technology Explained
According to a report by Nikkei Asia, Sony Group has developed a tool capable of identifying the original music that forms the basis of an AI-generated song. Since generative AI music models rely on vast datasets of songs and samples, the system is designed to trace the underlying copyrighted works used in the final output.
ALSO READ: India AI Impact Summit 2026: Government Expects Over $200 Billion Investment To Boost AI Adoption
The report claims that the technology can detect multiple original songs within a single AI-generated track. It can also reportedly indicate what percentage of each original composition contributed to the generated music. However, no technical details about how the system functions internally were disclosed.
Sony’s system reportedly works in two distinct ways. If AI developers cooperate, the tool can directly connect to their base models and analyse training datasets to identify which copyrighted works were used. In cases where developers do not cooperate, the system instead compares AI-generated music with existing music catalogues to estimate the original sources.
ALSO READ: Hackers Target Crypto Wallet Users With Fake Physical Letters To Steal Recovery Phrases
The development was led by Sony AI, the company’s research and development arm. A related academic paper describing the approach has reportedly been accepted at an international conference, indicating that the work has undergone peer review.
Sony reportedly envisions the technology being integrated directly into AI music models or used by rights holders during licensing negotiations. Although no commercial rollout timeline has been announced, the company sees it as a way to enable fairer revenue distribution for original songwriters.