The Polish data protection authority (UODO) organized an expert conference to discuss the legal issues raised by generative artificial intelligence, particularly with regard to the GDPR and copyright law.

During the event, the President of the UODO emphasized the urgency of these debates by citing a strategic report according to which 95.9% of the entities consulted do not consider themselves ready to implement AI in compliance with the GDPR. The keynote speaker, a doctor of law, then analyzed the challenges posed by generative AI, notably the difficulty in distinguishing real content from synthetic content and the false analogies created by the anthropomorphism of terms such as "learning" or "memory." He also highlighted the risk of "memorization," where a model can reproduce its training data. From a copyright perspective, legal uncertainty remains regarding the effective exercise of the right to object to the mining of texts and data for commercial purposes, and the transparency required by the AI law may prove insufficient for authors to verify the use of their works.

The subsequent panel discussion allowed for a deeper exploration of these topics, addressing the AI law from the perspective of businesses, the processing of special categories of data, the adaptation of copyright law, and the legal terminology to be used for AI.