Copyright, text & data mining and the innovation dimension of generative AI

Abstract

Abstract The rise of Generative AI has raised many questions from the perspective of copyright. From the lens of copyright and database rights, issues revolve not only around the authorship of AI-generated outputs, but also the very process that leads to the generation of these outputs, namely the process of text and data mining (TDM). Does unauthorized TDM process infringe the economic rights of the rightholders? How does the TDM-debate transform and transmute in the age of Generative AI? Generative AI tools create works that substitute the content creators whose very work that they learn from, and successively improvise themselves with every iteration. Generative AI, thus, also presents larger policy question as they substitute the romanticized human author that sits at the centre of copyright. In addition, as Generative AI tools, such as ChatGPT, can now also crawl the web, questions thus transcend the frontiers of copyright, and touch upon innovation and competition in the market for web browsers. This research article contemplates on the foregoing issues, and makes some recommendations to create a balanced framework, whereby incentives to innovate are preserved, and the interests of the human author are suitably safeguarded in the age of TDM and Generative AI.