The Evolving Landscape of Copyright: Lessons from the UK, China, and South Africa for Regulating AI-Generated Works in Ethiopia

Abstract

 The meteoric rise of artificial intelligence (AI), or more particularly the ability to
generate novel content, poses a powerful and new challenge to conventional
copyright. Traditionally, copyright has been founded upon the theory of human
authorship, an anthropocentric system that rewards and inspires human creativity.
But the modern generative AI systems are currently producing sophisticated works
without or with minimal human contribution, which leaves a legal gap and raises
fundamental questions as to who, if anybody, should be the owner of these works.
Ethiopia's existing legal regime, predominantly the Copyright and Neighbouring
Rights Protection Proclamation No. 410/2004 and its amendment Proclamation
No. 872/2014, is based on the human model and lacks any solitary special
provision concerning AI-generated works, which presents a fundamental obstacle
to their protection. In a bid to bridge this critical issue, this article provides a
comparative doctrinal and legal discussion of the distinctive approaches of the
United Kingdom, China, and South Africa. These jurisdictions offer useful
lessons: the UK's realistic approach of assigning authorship to the person who
takes the "necessary arrangements" in a computer-generated work; China's
innovative, case-by-case "human-in-the-loop" approach that grants copyright for
human creative input; and South Africa's potentially broad, yet so far untested,
legal foundation for computer-generated works. From this set of experiences, this
article derives some key recommendations for Ethiopia, including clarifying
authorship, creating a system of training data in AI, and establishing a flexible
legal and judicial infrastructure to respond to the complex future of copyright law
in the digital age.