Investigator:
Zihao Li

Abstract:
Artificial intelligence holds unprecedented potential to accelerate scientific discovery, yet researchers face persistent legal uncertainties when deploying AI tools in their work. Two areas are particularly salient: copyright, which governs the reuse of textual and data materials for training and analysis, and data privacy, which sets the boundaries for processing personal data in research contexts. While policymakers in the EU, UK, and US have sought to enable innovation through text and data mining (TDM) exceptions under UK and EU law, the fair use doctrine in the US, and research exemptions and evolving guidance within data protection frameworks, significant hurdles and uncertainties remain. These include doctrinal ambiguities, inconsistent scope of exceptions, heightened compliance costs, and chilling effects on interdisciplinary collaboration. This project offers a comparative analysis of the EU, UK and US approaches to copyright and data privacy in AI-driven research, identifying points of convergence, divergence, and potential regulatory blind spots. It argues that removing unnecessary legal hurdles is essential not only to unleash the potential of AI for science but also to ensure that research communities can pursue innovation responsibly, lawfully, and across borders.