In recent years, Generative Artificial Intelligence (GenAI) has revolutionized the production of texts, images, videos, and music. However, this innovation raises questions concerning copyright and online intellectual property, especially in Europe. This article offers a brief overview of the legal implications surrounding the use of protected content for training GenAI models, the European regulations on Text and Data Mining (TDM), and strategies for recognizing and tracking AI-generated digital content.
Index:
- How does generative AI work?
- What does the European legislation say?
- How to recognize AI-generated content?
- Not only challenges but also opportunities
- EUIPO’s Copyright Knowledge Centre
How does generative AI work?
Content generated by Generative Artificial Intelligence (GenAI) is created automatically by systems based on large language models and complex algorithms. These systems are “trained” on huge amounts of data: texts, images, videos, and audio. However, much of this content is proteceted by copyright, and is also acquired through web scraping, which is the automated extraction of online data, often without permission. This raises questions about online copyright protection, especially when the use is used for commercial or editorial purposes.
What does the European legislation say?
In Europe, the Copyright in the Digital Single Market Directive regulates the use of protected content for purposes of Text and Data Mining (TDM), an automated technique that analyzes text or numerical data to extract information, trends or patterns. Rights holders can object to the use of their content, through technical tools or legal notices in machine-readable format. It is therefore essential that those developing generative AI systems ensure that the data used for training are legally accessible and do not infringe on the copyright of digital works.
How to recognize AI-generated content?
One of the main regulatory concerns is theidentification of AI-generated content. The AI Act, the European regulation on artificial intelligence, introduces transparency obligations: artificially produced content must be recognizable through metadata or origin tracking systems. Specific technological tools, still under development, will help better manage this to protect online copyrights and provide legal safeguards for users and creators.
Not only challenges but also opportunities
Despite the risks associated with copyright infringements of digital works, new economic opportunities are emerging. Indeed, direct licensing agreements are emerging between copyright holders and generative AI developers to enable legal and remunerated use of content, especially in the editorial, musical and scientific sectors. To support this market, it will be crucial to have tools for mediation, monitoring, and institutional support, including software registration and licenses management.
EUIPO’s Copyright Knowledge Centre
To address the complexities associated with copyright and generative AI, by the end of 2025, the
EUIPO
will launch the Copyright Knowledge Centre with three key objectives:
- Inform holders on how to protect digital works;
- Provide monitoring tools on AI’s use of content;
- Fostering dialogue between developers, creators and legislators.
In summary, Generative Artificial Intelligence presents significant legal challenges for copyright, but also concrete opportunities to innovate the digital content market. In this context, the role of legal experts in intellectual property will be crucial to manage licensing, legal consultations and the responsible use of new technologies.