Regulatory asymmetries in EU scientific research law: data protection, copyright, and AI development

This article examines the relationship between data protection law and copyright law in the regulation of scientific research and the development of artificial intelligence (AI) in the European Union (EU). It argues that the coexistence of purpose-based and status-based regulatory models creates a fragmented legal environment for AI training and undermines the coherence of EU innovation policy. Data protection law adopts a purpose-oriented framework centred on scientific research and the public interest, whereas copyright law conditions research-related uses on institutional status and lawful access under the Digital Single Market (DSM) Directive. Through a doctrinal analysis of EU legislation, case law, and Estonian implementing law, the article demonstrates how this asymmetry generates legal uncertainty where AI training depends simultaneously on access to personal data and copyright-protected works. It further argues that the absence of a public-interest-oriented framework in copyright law limits its capacity to support scientific progress and technological innovation, particularly in smaller linguistic communities that rely on access to national-language datasets for AI development.