Source: Crowell & Moring

Crowell & Moring reports on a significant court decision that reinforces a fundamental principle of copyright law in the context of Artificial Intelligence. A federal court has affirmed that attributes of a person's voice—such as tone, pitch, or cadence—are not eligible for copyright protection. This ruling arose from a case where a plaintiff alleged that their voice was unlawfully used to train a generative AI system, arguing that the unique qualities of their voice constituted a copyrightable work. The court, however, dismissed this claim, reiterating the long-standing legal doctrine that copyright protects the expression of ideas, not the ideas themselves, and that a voice is a physical attribute or a tool for expression rather than a fixed, creative work.

The client alert emphasizes the importance of this decision for the burgeoning generative AI industry. The court's reasoning clarifies that while a specific recording of a voice can be copyrighted, the underlying vocal characteristics cannot. The decision distinguishes between the protectable sound recording and the non-protectable style or method of vocal delivery. This distinction is crucial as it sets a clear boundary for copyright claims related to AI voice cloning and synthesis technologies. The firm advises that while copyright may not be a viable avenue for protecting voice attributes, other legal frameworks, such as the right of publicity or unfair competition laws, may offer alternative means of recourse for individuals whose voices are used without authorization.