Παράκαμψη προς το κυρίως περιεχόμενο
European data
data.europa.eu
Η επίσημη πύλη ευρωπαϊκών δεδομένων

Navigating legal challenges in AI training: New ODECO report explores data openness and community rights

How legal frictions shape the future of data reuse and open web resources

The ODECO project, known for advancing research on open data ecosystems in Europe, has published a new report titled ‘Legal Frictions for Data Openness: Reflections from a Case-Study on Re-Use of the Open Web for AI Training’. The report delves into the complex legal and policy challenges involved in using open web data to train AI models, particularly large language models (LLMs) and generative AI systems.

Drawing on interviews with AI researchers, a dedicated online workshop, and a legal analysis of 41 relevant cases, the report highlights a critical tension: while the open web is often viewed as a digital common, its reuse for AI training frequently runs up against copyright, data protection, and broader regulatory frameworks. The report reveals that current efforts to make training datasets both legally and technically open are often insufficient, enabling well-resourced actors to extract data without meaningful contributions back to the commons.

A key theme is the need to strengthen community data sovereignty. The report explores alternative licensing models — such as the Nwulite Obodo LicenseKaitiakitanga Licenses, and OpenRAIL Licenses — that seek to balance openness with obligations on re-users, promoting fairer and more accountable data practices. It also advocates for a more nuanced understanding of legal ‘frictions’ not as barriers but as necessary checks that support healthier and more equitable data ecosystems.

By offering a deep dive into current legal controversies and emerging licensing strategies, this report provides valuable insights for anyone working with open data, AI, and digital policy. To learn more about ODECO’s broader work, revisit our earlier news piece on the ODECO project. For practitioners exploring licensing solutions, the Licensing Assistant on data.europa.eu can help you identify suitable licences for your data reuse needs, supporting transparent and responsible data practices.

For more news and events, follow us on X/Twitter and LinkedIn, or subscribe to our newsletter. You can also connect with other users through our collaboration channel.

Text of this article