Redesigning Copyright Protection in the Era of Artificial Intelligence

Abstract

How to protect the creations of artificial intelligence (“AI”) systems through copyright law is one of the most significant and complex legal issues of our time. Debated by stakeholders around the world, the issue has so far evaded consensus. However, a series of administrative and judicial rulings has altered the landscape. In 2019, a Chinese district court affirmed the copyrightability of a news article written by an AI-powered robot, clarifying that it was the AI developer that should be deemed copyright owner of the news article. In 2020, the United States Patent and Trademark Office, the European Patent Office, and the United Kingdom High Court prominently denied patent applications designating an AI system as an inventor.

This Article reveals the profound implications of these recent AI rulings for copyright protection of AI-generated works. It examines how these rulings shed new light on the nature and scope of authorship and ownership, two legal issues that are central to deciding whether and how AI-generated works merit copyright protection.

Drawing on its in-depth study of the recent AI rulings, the Article proposes a broad-based, forward-looking approach to protecting AI-generated works through a two-tiered legal mechanism. It suggests that legislators may consider awarding sui generis rights to AI works generated with human contributions, while those generated autonomously by AI systems without such contributions should be placed in the public domain without copyright protection.