Beyond Incentives: Copyright in the Age of Algorithmic Production

Abstract

Generative artificial intelligence (AI) art systems have, as with every prior artistic technological development, disrupted the status quo of the Copyright regime. Traditionally, the steep investment costs necessary to mass produce and distribute creative works dissuaded artists from profiting off their creations. After all, why would any artist undertake large scale capital investments if another could simply copy their work, undercut the original in the marketplace, and avoid the costs the original artist incurred? So the Copyright Incentive Theory argues that, without the Copyright privilege, the risk of free-riders disincentivizes artists from making and selling art. Copyright is designed as a monopoly privilege that offers increased (monopolistic) profits so artists are motivated to create new works. In turn, Congress should balance the resulting deadweight losses incurred through monopolization by adjusting various Copyright policy levers. Taken to an extreme, excessive Copyright protections conversely disincentivizes future creation by limiting the raw materials available for new works. This too results in deadweight losses through over monopolization.

The advent of generative AI upends these core assumptions. It calls into question the very foundations of American copyright law, and challenges the economic foundations of the utilitarian Copyright Incentive Theory. As technology evolves, long-held theoretical premises require re-examination against modern market realities. Facing an artistic singularity, we must assess how copyright can balance human and machine creativity for the public interest. Generative AI systems are capable of automating the entire creative process with minimal human involvement, superintendence, or labor. This upends the Copyright Incentive Theory’s careful economic balance, while exposing some of the myths that have emerged around the Copyright Incentive Theory. As the marginal cost of production of creativity trends towards zero, Copyright's premise of providing economic incentives for human creators becomes less relevant. As the risk of market substitution by machine authored works becomes more pronounced, Congress must reexamine the very premise of the Copyrights. It can do so by shifting its focus from incentivizing human creation to managing near post-scarcity abundance of content, enabling public access to digital works through a more robust system of digital rights management protection, and strengthening moral rights protections.

Keywords: Copyright, AI, Law and Economics, Copyright Incentives, Human Authorship, Machine Authorship