When AI builds on others' works

The new report, The Legal, Policy, and Market Impact of Artificial Intelligence on Text- and Image-Based Works, examines how generative AI is affecting copyright law and the economic conditions of the creative industries. The study was produced by the International Federation of Reproduction Rights Organisations (IFRRO) in collaboration with Eleonora Rosati, Professor of Intellectual Property Law at the Faculty of Law, Stockholm University.

Through its member organisations, IFRRO represents millions of authors, visual artists and publishers worldwide. The report builds on the organisation's previous work on AI and copyright and analyses the legal, market and policy implications of generative AI for text- and image-based works. It is also the latest publication produced within the framework of IFRRO's ongoing collaboration with Eleonora Rosati.

Particular attention is given to the impact on rightholders when their works are used for AI training, and to how legislators, courts and policymakers may respond to these developments.

Revenue losses and a growing licensing market

More than 9 respondents out of 10 agree that genAI has already negatively impacted their earnings.

Source: IFRRO, The Legal, Policy, and Market Impact of Artificial Intelligence on Text- and Image-Based Works (2026).

As part of the study, an international survey was conducted among IFRRO member organisations. The findings show that generative AI has already resulted in measurable revenue losses for certain categories of rightholders. At the same time, a market for licensing content for AI training purposes has emerged and continues to expand.

“All this underlines the need for a sustainable AI ecosystem in which rightholders can control and consent to the use of their works, copyright principles are upheld, and transparency is ensured,” says Eleonora Rosati.

Rosati's research cited in the first AI copyright case before the CJEU

These issues are also playing an increasingly important role in EU legal developments. In November 2026, Advocate General Maciej Szpunar delivered his Opinion in the Like Company case before the Court of Justice of the European Union, the first generative AI case to reach the Court.

The case concerns Google's AI service Gemini and its use of journalistic content. A Hungarian publisher alleges that the service uses material from its websites to generate responses for users without authorisation or remuneration. The Court has therefore been asked to consider under what circumstances AI services may infringe neighbouring rights held by press publishers.

In his Opinion, the Advocate General refers to three of Eleonora Rosati's publications on copyright and AI. These works address, among other issues, the EU framework for text and data mining, the role of copyright within the Digital Single Market, and liability questions relating to AI-generated outputs. They are used to provide context for the EU legal framework and the broader relationship between copyright and artificial intelligence.

“The Advocate General addresses several of the most important issues currently arising at the intersection of AI and copyright. At the same time, he takes the view that the Court should not answer the questions concerning AI training in this particular case and stresses that one should not draw overly broad conclusions from the Opinion,” says Eleonora Rosati.

A rapidly evolving field

The Advocate General's Opinion is not binding on the Court of Justice, which will deliver its judgment at a later stage. Together with IFRRO's report, however, it illustrates how copyright issues have become central to discussions about the future development of generative AI.

For Eleonora Rosati, it is essential that technological progress does not come at the expense of the individuals and organisations whose content is used to develop AI systems.

The report therefore highlights the importance of consent, transparency and effective copyright protection, while the ongoing Like Company proceedings demonstrate how these issues are now being tested in practice. Together, they illustrate how central copyright has become to the future development of AI and how research at the Faculty of Law, Stockholm University, contributes to the international discussion on the legal implications of emerging technologies.

Text: Natalie Oliwsson