Man, Machine, or Both? Reconstructing AI-Patententability under the Patents and Designs Act, 2004

Abstract

Artificial intelligence has transform technological innovation with machine learning, and neural networks, with generative AI increasingly participating in conception, testing, optimisation, and implementation of product and processes. This development has added pressure to the patent systems designed around human inventors and technologies classifiable when the Patents and Designs Act, Cap P2 LFN, 2004 was enacted. This article interrogates the patentability of AI-related inventions within Nigeria's patent framework. It argues that AI should not constitute a legal barrier to patentability, but that, the proper inquiry should remain whether claimed subject matter satisfies the statutory requirements of novelty, inventive activity, and industrial application under section 1 of the Act, while the law separately determines human inventorship and ownership consequences under sections 2 and 3. The article draws on enacted legislations, case laws, scholarly writing and comparative study to illuminate the subject matter. The paper proposes technological-neutral reforms for reconstructing the intersection between AI patentability and the Patents and Designs Act. These reforms include: recognition of AI-assisted inventions to a define limit under the Act; clarifying the position on computer-software and implemented inventions; establishing rules for AI-assisted inventorship and disclosure; improving substantive examination capacity, and creating a coherent framework for data-intensive innovation. The paper concludes that Nigeria can use patent law not merely to accommodate AI but to stimulate domestic research, technology transfer, commercialisation, and industrial development.