In May 2025, the European Union Intellectual Property Office (EUIPO) released a status report about the development of generative AI and Large Language Models.
Copyright-intensive sectors alone account for more than 17 million jobs and nearly 7% of the EU’s GDP, underlining the central role of intellectual property in driving Europe’s prosperity and global competitiveness.
Over the past three decades, successive waves of digital innovation have reshaped the way content is created, distributed and accessed. Throughout these transformations, copyright law has adapted to ensure that creators receive recognition and remuneration for their work, thereby sustaining the creative sectors that enrich our societies.
Why this study?
The main objective of the study is to analyse copyright implications at both the input and output stages of GenAI, focusing on the related technical solutions:
- On the input side, the purpose is to analyse technical solutions and practices currently used, or still under development, to reserve, limit or licence the use of copyright protected works as training material for the development of GenAI systems.
- On the output side, the purpose is to analyse technical solutions and practices to identify content generated by, or with the help of GenAI, as well as practices to prevent the generation of content that might infringe on existing intellectual property rights. The central premise of this study is thus a ‘solution-driven approach’.
Generative AI Ecosystem
The report defines a ‘Generative AI Ecosystem’ comprised of legal and commercial stakeholders in the development and use of GenAI technologies, as well as the legal, commercial, and institutional relationships that arise from their interactions.
The top layer (‘EU Legal Chain’) lists the entities that form part of the value chain, based on the relevant EU legal provisions applying to their activities, including Copyright owners, users of text and data mining (TDM) features, AI syste providers, deployers and end-users.
The middle section (‘Value Chain’) details the different categories of actors throughout the value chain falling under categories defined in the top layer including copyright owners and their intermediaries and model developers.
The bottom layer of the diagram (‘Types of Measures’) contextualizes where measures are applied in the value chain
Main findings
The study shows that:
- No single solution has emerged as the sole standard opt-out mechanism for rights holders to express their TDM rights reservations, or transparency measure to identify and disclose the nature of synthetic content.
- Although the global GenAI landscape involves a rising number of legal disputes, the study also notes that several high-value agreements have been reached between rights holders and GenAI developers.
- The current situation suggests a possible role for public authorities in providing technical support for implementing and administering databases of TDM reservations and raising awareness on measures and good practices to mitigate potential infringing output.
As a disrupting technology, the development of GenAI has caused shifts in the creative and the IT industries, and significantly altered how rights holders and AI developers operate. While it may take some time before a new balance is established, the study importantly showed the relevance of accessing essential information about works’ origin and permissible uses in view of proper respect, benefit and enforcement of copyright.
Research methodology
This study has been prepared by a research team of the University of Turin Law School and the Nexa Center for Internet & Society from the Polytechnic of Turin for the European Union Intellectual Property Office (EUIPO). The research that took place between September 2024 and March 2025 was supported by extensive desk research as well as interviews from key stakeholder groups including, copyright holders, AI companies, technology solution providers as well as public organisations.
Further reading
Development of Generative Artificial Intelligence from a Copyright Perspective. Landing page with links to report and executive summary. May 15, 2025. European Union Intellectual Property Office (EUIPO)
Why it matters
This study is designed to clarify how GenAI systems interact with copyright – technically, legally, and economically. It examines how copyright-protected content is used in training models, what the applicable EU legal framework is, how creators can reserve their rights through opt-out mechanisms, and what technologies exist to mark or identify AI-generated outputs.
It also explores licensing opportunities and the potential emergence of a functioning market for AI training data. Although the study is intended for experts in the field, it lays the groundwork for developing clear and accessible informational resources for a broader audience.
