HUMAN-REQUIRED ORIGINALITY: RECALIBRATING COPYRIGHT FOR A POST-AI WORLD

Generative artificial intelligence upends the assumptions that have anchored U.S. copyright law for more than a century. By enabling the production of high‑quality expressive works at effectively zero marginal cost, GenAI destabilizes copyright’s utilitarian foundation and exposes a deep incoherence in the originality requirement: the law grants exclusive rights to human‑authored works that could have been produced just as easily—and at no cost—by modern generative systems, while denying protection to the machine‑generated equivalents that render human effort unnecessary. At the same time, GenAI introduces a dynamic risk that existing doctrine is unequipped to address: as AI increasingly substitutes for human creators in markets characterized by commoditized content, human participation may collapse, starving future models of the novel, high‑signal training data needed for continued aesthetic and cultural evolution. Left unaddressed, these twin pressures threaten both the theoretical coherence and the long‑term creative vitality of the copyright system. This Article argues that the core of the problem lies not in questions of infringement or AI authorship, but in copyright’s threshold requirement: the “modicum of creativity” standard no longer filters works that require human incentives from those that do not. I propose a new, technologically grounded interpretation of originality—the human‑required creativity standard—under which copyright protection attaches only to works that could not have been generated by state‑of‑the‑art models with de minimis human input at the time of authorship or registration. This content‑focused approach restores alignment between copyright’s incentives and its constitutional purpose by denying protection to trivially generable works regardless of whether a human or a machine produced them, while continuing to protect works that demand meaningful human creative contribution. A rebuttable presumption of eligibility and an affirmative defense of trivial generability make the human required standard administratively workable while preserving automatic copyright protection. The framework remains compatible with international treaty obligations, avoids prohibited formalities, and maps onto familiar originality and derivative‑work doctrines. By grounding copyright eligibility in the realities of modern creative production, the human‑required standard offers a path toward preserving human creativity, supporting sustainable innovation in AI, and re‑anchoring copyright law in its fundamental task: promoting progress in the arts and sciences.