As artificial intelligence (AI) continues to transform the ways in which content is accessed, processed, and used, the protection of copyright-protected works in the context of text and data mining (TDM) has become an increasingly complex and urgent issue. Content creators and rightsholders face significant challenges in ensuring that their opt-outs from TDM activities are clearly visible, enforceable, and respected—especially as AI developers mine vast datasets to train their systems.
To address these emerging needs, Visionary Analytics, partnering with Capgemini Invent is conducting an EU-funded study by the European Commission (DG CONNECT) to assess the policy, technical, and practical implications of creating a centralised work-based registry of TDM opt-outs. Such a registry would allow rightsholders to express reservations on the use of their works for AI training and other purposes, providing AI developers with a clear, authoritative source of rights information. This mechanism could streamline compliance, improve legal certainty, and reduce the risk of unintentional infringement.
The study will explore five key dimensions:
-
Mapping the landscape of existing TDM opt-out solutions and identifying critical gaps.
-
Assessing the need for and potential value of a centralised registry for all stakeholders.
-
Evaluating the technical and operational feasibility of such a registry.
-
Analysing the essential features and technical requirements needed to make the registry effective and sustainable.
-
Designing a governance framework and long-term strategy to ensure the registry’s viability.
Visionary Analytics will apply desk research, literature review, mapping of existing opt-out mechanisms, and stakeholder interviews to inform these findings. By producing evidence-based recommendations, this project aims to support future EU-level solutions that balance the interests of rightsholders, AI developers, and society at large.
Stay tuned for insights from this important initiative at the intersection of copyright, technology, and AI policy.
The project is expected to be finished by February 2026.
