Do LLMs 'store' personal data? This is asking the wrong question

MoFo’s Lokke Moerel and Marijn Storm published an article with IAPP discussing the merits of the Hamburg and Danish data protection authorities’ position that LLMs do not contain personal data and therefore (i) data subjects’ rights cannot relate to the model itself and (ii) models trained in violation of the EU General Data Protection Regulation do not affect the lawfulness of the use of such models within an AI system. Read the full article.

The article is the third in a series. Previous articles:

• Using special categories of data for training LLMs: never allowed?, discusses the merits of the interim report by the EDPB’s ChatGPT Taskforce stating that LLMs may not be trained with special categories of data
• Data Subjects as Controllers: Violation of GDPR's Fairness Principle, discusses whether individuals can be considered controllers for their own inputs and outputs, when using LLM chatbots.