The Dilemma of False Positives: Making Content ID Algorithms more Conducive to Fostering Innovative Fair Use in Music Creation
Published Web Locationhttps://doi.org/10.5070/LR8241035525
Content ID programs commonly use algorithms to block uploaded music when the algorithm concludes the owners of certain copyrighted works will claim their work is being used without consent. However, algorithmic enforcement programs can produce “false positives,” where legally allowable music associated with a reference file is inappropriately blocked. The phenomenon of false positives is especially problematic for songwriters, composers, experimental music artists and others who create music by combining their own vocal or instrumental performance with work created by others and “loops” from audio libraries. Balanced by such factors as how much a new work damages the market for a prior work and how much of a prior work is used in a new work, the “fair use” defense allows songwriters to upload technically infringing work if the new work amounts to a critique, is in the public domain, or sufficiently transforms the original work to render it new. This article explains how Content ID algorithms are developed and interpreted and discusses how the fair use defense can sometimes limit the extent to which Content ID programs can block innovative music creation. The article offers methods for defining and measuring algorithmic effectiveness that both account for the risk of false positives and protect the proprietary interests of copyright holders. It also proposes a new regulatory scheme that ensures these methods are implemented properly. The proposed regulatory scheme should lead to a more equitable system for music creators and original copyright holders and to more inventive and interesting music for fans.