Kim on Authorship in the Age of AI

So Yeon Kim (UC Berkeley School of Law) has posted Brushing The Contour Of Authorship In The Age Of AI: A Moral Rights And Comparative Law Framework (University of Pennsylvania Asian Law Review) on SSRN.  Here is the abstract:

Generative artificial intelligence (AI) unsettles a longstanding premise of U.S. copyright law: that authorship is inherently human. This Article explores the philosophical justification for granting copyright to human authors. It advances a human-centered account of authorship for U.S. copyright law through John Locke’s authorial rights framework, informed by comparative moral-rights analysis. It also draws implications from a less explored East Asian, Confucian-inflected understanding of moral rights for AI-era disputes. Avoiding what some call the “AI copyright trap”—the assumption that existing copyright law can govern generative AI issues—this Article reframes authorship as an intellectual and personal connection between creator and work.

This Article argues that current U.S. copyright law emphasizes human authorship and intention, but has not fully explained why the human connection to the work matters normatively. To address that gap, I analyze Hegel’s and Kant’s personhood theories alongside Locke’s labor and authorial rights frameworks. I argue that Hegelian theory falls short because it begins from the premise that expressive works are normatively distinct from other objects, when that distinctiveness is better understood as arising from their unique connection to authors.

I then distinguish between appropriation and possession or enjoyment. The question of why the law should grant exclusionary rights is different from the question how far those rights should extend over later uses. That distinction matters in copyright, where expressive works are non-rivalrous and often cumulative or transformative. For this reason, I argue that Mala Chatterjee’s Lockean authorial rights framework offers a more coherent explanation than personhood theories. It better distinguishes between mere use and wrongful appropriation, and better accounts for independent creation, transformative use, fair use, and AI-assisted creation.

This Article then analyzes moral rights infringement and adjacent cases both from the U.S. and Korea, reflecting recurring problems of authorship in the age of AI. I argue that the Lockean authorial rights framework helps clarify the human connection to the work while remaining within copyright’s traditional limits. At the same time, the framework alone cannot fully explain harms such as misattribution, distortion, or AI-enabled omission of human contribution. Comparative analysis therefore becomes essential. The Lockean authorial rights framework helps explain why creative labor merits protection, while Confucian understandings of authorship help explain dignity-based harms independent of economic injury. I further show that, despite formal divergence, both the U.S. and Korean systems already balance authors’ moral interests with social and economic concerns. Accordingly, I propose a renewed framework for the age of AI, centering the human author’s connection to the work through Lockean authorial rights theory, while drawing on moral-rights-inflected and Confucian insights where dignitary interests are implicated.

To receive new posts from Legal Theory Blog by email, get a free subscription to Legal Theory Stack.

Lawrence Solum