As every year, the Department of Law and Technology at the Faculty of Law, in cooperation with the Faculty of Social Studies, organized a two-day conference, from the 29th to the 30th of November, to bring together academics, practitioners, and students for an interdisciplinary exchange on the legal and societal implications of information and communication technologies.
The ID Law department took part in the track “Legal dimensions of the use of synthetic data in research” chaired by Prof. Nikolaus Forgó.
The presentation held by Kseniia Guliaeva, Ibrahim Sabra (both representing Phase IV AI) and Olga Startseva (representing SYNTHIA) examined the legal status of synthetic data as personal or non-personal data, and its legal status in healthcare AI under the GDPR. Additionally, they explored the legal classification of generation and use of synthetic data under the GDPR. The presentation draws on two EU-funded healthcare projects. Phase IV AI aims to create privacy-preserving synthetic data generation and secure multi-party access to address health data challenges for AI. Phase IV AI will provide anonymised datasets, build a Data Market to promote sharing of data, and incorporate these into a cross-European health data hub.
The legal implications of determining whether synthetic data qualifies as personal data are highly significant for SYNTHIA. The UNIVIE team is analysing the legal standards surrounding re-identifiability, specifically, the level of anonymity synthetic data must achieve to remain compliant with privacy regulations. Our work informs how synthetic datasets can be developed and deployed safely. These legal insights are essential as SYNTHIA moves forward with its core outputs. The project will deliver synthetic datasets across six disease areas and demonstrate their value in both research and clinical applications. The results of the project will be made openly accessible to promote transparency and broad adoption.
Hannah Hafenscher, Selen Yakar and Wolfgang Alexander Kolar represented the FFG-Project „Prevention through digitisation and standardisation of self-reported data” („PräNUDGE“) and delivered a presentation on whether synthetic data could serve as a “cure” for de-biasing algorithmic systems in medical research. The talk focused on the underrepresentation of certain groups and explored how synthetic data can be used to address demographic imbalances in real-world datasets by highlighting the relevant legal dimensions under the GDPR and the AI Act.
Michael Schmidbauer, representing the FFG-Project "AI-based Planning and monitoring of buildings: towards Data sovereignty" (plAIdata), held a presentation titled “Synthetic Data in the EU: IP Perspectives and Data Sovereignty”. He talked about the (lack of a) European legal framework for generation and protection of synthetic data from an IP perspective and what this means for the broader dependability on rare, often non-European data sets for AI training.
Klaudia Kwiatkowska and Alexandra Mărginean, representing eBRAIN-Health, addressed in their presentation the little attention that synthetic data has received under the European Data Strategy (more specifically under the Data Governance Act, Data Act and European Health Data Space), despite of its growing significance in research and in practice. One example of the advantages in using synthetic data in medical research is the eBRAIN-Health Project, where synthetic data will be used to enrich machine learning and deep learning algorithms which will support early diagnosis of neurodegenerative diseases.
We are grateful to Masaryk University for the opportunity to take part in this highly engaging conference and to share our research.
