Yang Chen (Law) has posted “Two Roads Diverge and Converge in the AI Era: Computer-Generated Works as an Exception for Human Authorship?” (36 Columbia Journal of Asian Law (forthcoming)) on SSRN. Here is the abstract:
The intersection of copyright law and artificial intelligence (AI) presents numerous unresolved challenges, while the copyrightability of AI-generated content (AIGC) is particularly complex. Jurisdictions have adopted distinct approaches that converge and diverge in addressing this matter. Classic human authorship jurisdictions, such as the United States (US), China, and the European Union (EU), require AIGC to satisfy human authorship and originality requirements to be copyrightable. As a result, their debates focus on the level of original authorial contributions required for AIGC protection. In contrast, jurisdictions following the computer-generated works (CGWs) doctrine, such as the United Kingdom (UK) and Hong Kong (HK), diverge significantly when applying a plain interpretation approach. These jurisdictions treat AIGC as copyrightable subject matter regardless of originality standard, even though they also generally follow the human authorship requirement. The convergences and divergences among these jurisdictions make their comparison particularly valuable to explore.
To unpack this convergence and divergence, the article first examines relevant scholarly debates and recent legal developments in classic human authorship jurisdictions (represented by the US, China, and the EU) that illustrate how they address the copyrightability of AIGC. It then explores the CGWs doctrine in jurisdictions such as the UK and HK, highlighting how their approaches may differ significantly. In light of these divergences, particularly under a plain interpretation approach, the article argues that the divergent stance adopted by CGWs jurisdictions is doctrinally problematic and lacks theoretical justification. The article proposes a refined doctrinal approach to the CGWs doctrine that aligns more closely with the frameworks in human authorship jurisdictions regarding AIGC protection. While some divergences may persist, this refined approach establishes a more reasonable and justifiable framework, fostering greater convergence between the two systems.
This comparative analysis offers insights on unified pathways forward for jurisdictions navigating AI and copyright. These pathways include developing a more reasonable originality standard for AIGC, exploring limitations on AIGC copyright protection, and acknowledging the diminishing role of copyright law in the AI era. This recognition of the diminishing role underscores the need for alternative solutions beyond the traditional copyright regime to address emerging challenges more directly and efficiently.
