AI Replication of Musical Styles Points the Way to An Exclusive Rights Regime

Abstract

While “style” has been considered at the margins of copyright, it may be central to what makes a work original and protectable. At the same time, an artist’s style as established over multiple works seems even more problematic for copyright’s core fixation requirement. Other misappropriation regimes such as rights of publicity or Lanham Act false affiliation/endorsement offer limited avenues to protection of style. Yet, style standing separate from any particular work has clear aesthetic and commercial value. The success of current AI music generators to use machine learning to create an algorithmic engine capable of producing musical works, performances, and recordings recognizably in the style of famous composers and performers indicates that style standing alone can be quantified and fixed. This chapter thus argues that musical style should be protectable on multiple grounds, including, if nothing else, unjust enrichment of tech companies at the expense of human artists.

Keywords: style, copyright, music, musical style, artificial intelligence, AI, machine learning, right of publicity, Lanham Act, affiliation, endorsement, exclusive rights, aesthetic value, commercial value