Article 102 TFEU and privacy harms in generative AI: A case for exploitative theories of harm

Abstract

This article examines whether privacy-related harms arising from generative artificial intelligence (GenAI) can constitute exploitative abuses under Article 102 TFEU. As GenAI systems increasingly rely on large-scale data collection and processing, concerns about user autonomy, consent, and control over personal data have become central not only to data protection law but also to the functioning of digital markets. The article argues that, in certain circumstances, the extraction and use of personal data by dominant firms may amount to unfair trading conditions within the meaning of Article 102 TFEU. Rejecting both strict separation and full integration of competition and data protection law, the article advances a polycentric approach, under which privacy considerations may be incorporated into competition law analysis where they affect market structure or the competitive process. Drawing on recent case law, including Meta Platforms, and examples from AI-driven markets, it demonstrates how privacy infringements can reinforce market power and distort competition. The article concludes that recognising privacy exploitation as a form of abuse provides a necessary and coherent basis for addressing the competitive implications of data practices in the GenAI ecosystem.