Sony AI Develops Technology to Analyze "Inspiration Sources" of AI-Generated Music
Sony's AI research division has introduced a technology capable of analyzing music tracks created using artificial intelligence and identifying the real compositions that served as their inspiration.
The key feature of this development is not merely detecting similarity between generated and original music, but also quantifying the contribution of specific works to the final AI track. For example, the system can determine that a certain AI track is 10% inspired by Queen's work and 30% by The Beatles.
Such detailed comparison creates legal preconditions: in the future, rights holders of original compositions could potentially make claims against the authors of AI-generated content.
The technology operates in two modes.
- In the first case, if the developer of the AI generator is willing to cooperate, Sony gains access to data by connecting to the base model. This allows for direct analysis of the training material datasets and precise tracing of which specific works were used to train the neural network.
- In the second case, when the creator of the music generator refuses to provide data, the system works differently: it analyzes only the finished AI track. The program then compares it with an extensive database of existing compositions and assesses probable sources of influence, identifying potential "prototypes" among real musical works.
Sony views this development as a foundation for a future system of revenue distribution from AI-generated content. According to the company's vision, the technology will enable accounting for the contribution of original work authors when monetizing generated tracks.
At present, representatives of AI music generator developers have not publicly commented on the initiative.