The European Parliament has just published a study commissioned by its Policy Department for Justice, Civil Liberties and Institutional Affairs which analyses the topic of ‘Technological Aspects of Generative AI in the Context of Copyright’.
This in-depth analysis explains the statistical nature of generative AI and how training on copyright-protected data results in persistent functional dependencies with respect to the used data. It highlights the challenges of attribution and novelty detection in these high-dimensional models, emphasizing the limitations of current methodologies. The study provides technical recommendations for traceability and output assessment mechanisms. This study is commissioned by the European Parliament’s Policy Department for Justice, Civil Liberties and Institutional Affairs at the request of the Committee on Legal Affairs.
In conclusion, the current capabilities of generative artificial intelligence (GenAI) models are grounded in statistical approximation and high-dimensional function learning. These systems do not operate through inexplicable means but rely on probabilistic interpolations derived from vast training datasets. While the internal representations they build, often in the form of complex hypersurfaces, are not directly interpretable, they are ultimately governed by mathematical principles, and thus susceptible to systematic evaluation, refinement, and regulation.
https://www.europarl.europa.eu/thinktank/en/document/IUST_BRI(2025)776529