The Work of Copyright Law in the Age of Generative AI
Kate Crawford, Jason Schultz
Headlines about generative artificial intelligence (AI) have forecast a coming collapse of the art world, the labor market, and even human sovereignty itself. But a more likely and imminent outcome is a profound destabilization of copyright law. For over three hundred years, Anglo-American copyright law has provided a legal backdrop for the arts and creative industries, a substrate of rules and regulations to govern economic and, to a lesser degree, social and cultural norms of how and when art can be made, sold, copied, and kept. In prior centuries, when new technologies such as the photograph, the sound recording, the computer processor, and the data server emerged, copyright adapted, creating odd and sometimes contorted doctrines to account for the epistemic challenges that each technology posed to the questions What is art? and Who creates it? Now the tables have turned, presenting more of an epistemic crisis of law. We see a latent but profound undermining of several foundational concepts in copyright itself. Generative AI attacks copyright at three major pressure points—authorship, expressiveness, and agency—raising the question of what copyright law will look like and what work it will do in this new era of computational production.