[Paper Review] Originality and the Future of Copyright in an Age of Generative AI

Paulius Jurčys, Mark Fenwick|arXiv (Cornell University)|Sep 8, 2023

Law, AI, and Intellectual PropertyComputer Science3 citations

TL;DR

This paper examines the legal and ethical implications of generative AI on copyright law, focusing on the concept of human authorship in AI-generated works. It argues that current copyright frameworks must evolve to address the challenges posed by AI systems that produce original content without direct human creative input, proposing a reevaluation of originality standards in intellectual property law.

ABSTRACT

This papers explores the question of human authorship when works are created with generative AI tools.

Motivation & Objective

• To analyze the tension between AI-generated content and existing copyright doctrines centered on human authorship.
• To investigate whether works produced by generative AI tools can satisfy the legal threshold of originality under current copyright law.
• To assess the implications of AI authorship for intellectual property rights, innovation, and legal accountability.
• To propose a framework for redefining originality and authorship in light of advances in generative AI technologies.
• To guide policymakers and legal scholars in adapting copyright law to the realities of AI-driven content creation.

Proposed method

• Conducts a doctrinal and comparative analysis of copyright law across major jurisdictions, focusing on originality requirements.
• Examines landmark court decisions and statutory frameworks to identify gaps in legal recognition of AI-generated works.
• Analyzes the role of human involvement in AI-assisted creation processes to determine the threshold for human authorship.
• Uses conceptual and normative legal reasoning to assess whether AI-generated content can be considered 'original' under existing standards.
• Proposes a revised legal framework that distinguishes between human-created and AI-generated works based on the degree and nature of human creative input.
• Engages with philosophical and legal theories of authorship and originality to inform policy recommendations.

Experimental results

Research questions

• RQ1 To what extent can works generated by AI systems meet the legal threshold of originality under current copyright law?
• RQ2 What constitutes sufficient human authorship when AI tools are used in the creative process?
• RQ3 How should copyright law adapt to recognize or exclude AI-generated content from protection?
• RQ4 What are the legal and ethical consequences of granting or denying copyright protection to AI-generated works?
• RQ5 How can legal systems balance innovation incentives with the need for clear authorship and ownership rules in the age of generative AI?

Key findings

• Current copyright law's emphasis on human authorship creates a significant barrier to recognizing AI-generated works as protectable.
• The threshold of originality is increasingly challenged by AI systems that produce complex, novel outputs without direct human intervention.
• The paper finds that existing legal frameworks are ill-equipped to handle cases where AI tools autonomously generate content that meets traditional originality criteria.
• There is a growing consensus that a new legal category—potentially for 'AI-assisted' or 'AI-generated' works—may be necessary to address the gap in protection.
• The study concludes that without legal reform, uncertainty in ownership, licensing, and infringement will hinder innovation and investment in AI-driven creative industries.
• The authors advocate for a functional approach to originality that evaluates the role of human input rather than requiring human creation as a prerequisite for protection.