Last week, several major record labels filed copyright infringement lawsuits in US courts against the makers of two generative AI music apps, Suno and Udio. The labels allege the AI companies have engaged in copyright infringement by copying many sound recordings belonging to the record labels, and producing outputs very similar to those recordings.
These lawsuits are not the first to trouble the booming generative AI industry. Visual artists have sued makers of image generating systems, while various newspapers are suing OpenAI, the owner of ChatGPT, for similar allegations. The result of the litigation may determine the future viability of such generative AI products.
To generate the song, the AI has been trained with a vast amount of data. The lawsuits allege this data comprises pre-existing sound recordings owned by various record labels and copied without permission. These sound recordings are at the heart of this issue. How will the court make its decision? The court will examine four factors in relation to the use of the record labels’ songs by Suno and Udio. These are:
The court will examine the amount and substantiality of the portion of songs copied. It will examine how the allegedly copied songs are used in the AI training process and in generating output. What’s in a voice? One major concern for the music industry is the cloning of voices. This is where other generative AI music apps can be used to clone a famous singer’s voice onto any song.