Watermarking Training Data of Music Generation Models
This addresses the problem of copyright infringement in AI-generated music for content creators and legal stakeholders, though it is incremental as it builds on existing watermarking methods.
The study investigated whether audio watermarking techniques can detect unauthorized use of copyrighted content in training music generation models, finding that watermarks, including imperceptible ones, cause noticeable shifts in model outputs and are robust to removal techniques.
Generative Artificial Intelligence (Gen-AI) models are increasingly used to produce content across domains, including text, images, and audio. While these models represent a major technical breakthrough, they gain their generative capabilities from being trained on enormous amounts of human-generated content, which often includes copyrighted material. In this work, we investigate whether audio watermarking techniques can be used to detect an unauthorized usage of content to train a music generation model. We compare outputs generated by a model trained on watermarked data to a model trained on non-watermarked data. We study factors that impact the model's generation behaviour: the watermarking technique, the proportion of watermarked samples in the training set, and the robustness of the watermarking technique against the model's tokenizer. Our results show that audio watermarking techniques, including some that are imperceptible to humans, can lead to noticeable shifts in the model's outputs. We also study the robustness of a state-of-the-art watermarking technique to removal techniques.