Abstract
The fair use limitation on the scope of authorial exclusive rights is expected to be the main defense in lawsuits charging generative artificial intelligence (AI) developers with copyright infringement. These lawsuits—brought by and on behalf of some authors, artists, and copyright industry rights holders—challenge the legality of the unauthorized use of in-copyright works for purposes of developing generative AI systems, which they allege has greatly harmed the plaintiffs. The outcome of these copyright infringement suits will likely turn on courts’ assessments of the fair use doctrine’s market effects factor: whether and to what extent generative AI is harming or is likely to harm plaintiffs’ existing and expected markets.
Controversies about new and unexpected uses of copyrighted works enabled by technological advances are far from new; they continue to play an integral part in the development of modern copyright law. This Article revisits numerous disputes over uses of in-copyright works enabled by disruptive technologies since the 1970s. These disputes involved copying of research and educational materials, home video and audio taping, and storage and processing of digital copies for the purpose of developing search engines and other research tools. Tracing the evolution of the concept of market harm and the role it has played in copyright fair use cases involving new technologies, this Article lays out the arguments about market effects—harmful and beneficial effects, as well as the requirements of proof—that have been presented in the courts and debated in Congress. The Article then assesses the current landscape of generative AI copyright litigation. Insights gained from the past can be instrumental in thinking about conflicts involving generative AI, as well as any future technologies that are new, innovative, and disruptive.
