The intersection of Intellectual Property (IP) and Artificial Intelligence (AI) presents complex challenges that test the foundational assumptions underlying authorship, creativity and innovation. Legal frameworks, such as copyright and patents, grant authors and innovators exclusive rights, preventing others from profiting from their work without consent. These policy tools aim to enable and sustain fair competition among creators. Yet, the advent of Generative AI places creative labour in a state of precarity and, therefore, risks perpetuating economic injustices, especially for African artists.
Responding to these challenges, this paper reflects on IP’s original intentions: to protect human creativity and provide incentive for innovation. It argues that while AI may demonstrate some creative workflows, its investments and innovation incentives are ultimately not impacted by IP protection, whereas the incentives of artistic economies are. By facilitating easy access to low-cost reproductions of creative outputs, using copyright works as its training data, GenAI threatens the demand for human work, lowering the cost, quality and diversity of representation in the creative sector.
As this piece demonstrates, multiple infringement cases illustrate the extent of disruption caused by AI in the creative industries. However, AI service providers continue to claim ‘fair use’ or ‘fair dealing’ to exempt themselves from compensating thousands of creatives. Without clear legal precedent, exceptions for AI training or outputs, and a lack of distinction between human and automated creativity, litigation becomes difficult. Beyond copyright, similar disruptions also exist in patent systems, where AI is used to both produce patent applications and examine patent claims.
Through AI’s ability to access the entire patent claims corpus at once and produce logical applications with ease, arguments have arisen regarding AI’s status as an inventor. In South Africa, Dabus AI was granted status as a co-inventor for its patent; elsewhere around the world, these claims have been denied. In both copyright and patents, clarity is needed regarding AI’s status as an author and inventor. To thoroughly answer some of these questions, scrutiny over AI algorithms and clear disclosure surrounding AI-generated work are imperative. However, with service providers beginning to use Trade Secret protection to conceal training sets and AI interaction history, regulation becomes difficult.
Using IP as a key subject of inquiry to assess economic justice in Africa and in AI governance, the paper calls for IP legal reforms that can address the specific needs and contexts of African countries, balancing IP monopolies with public-interest considerations.
