Michèle Finck has just published a new article on the research exemptions in the EU Artificial Intelligence Act. This piece explores one of the most complex and under-examined aspects of the EU AI Act — its research exemptions. These provisions were designed to ensure that AI research and innovation can flourish without excessive regulatory burdens. However, as I argue in the article, the exemptions are ambiguous and difficult to apply in practice, leaving researchers uncertain about when their work falls within or outside the Act’s scope.

This means that the “lost research exemption” may not provide the safe space for research that the legislators envisioned. Michèle call for clearer interpretive guidance to ensure the AI Act truly supports innovation while maintaining accountability.

You can read the article in open access here.

The Chair for Law and Artificial Intelligence hosted a Writing Workshop on the European AI Act in Tübingen, Germany, on 11 July 2025.

The article “Robustness and Cybersecurity in the EU Artificial Intelligence Act“ by Henrik Nolte, Miriam Rateike, and Professor Michèle Finck has been published on arXiv. It examines the legal challenges and shortcomings in the provisions related to robustness and cybersecurity for high-risk AI systems (Article 15 AIA) and general-purpose AI models (Article 55 AIA) in the EU Artificial Intelligence Act.

The Chair of Law and Artificial Intelligence at the University of Tübingen hosted its second International PhD Summer School in Artificial Intelligence and EU Law from 12 – 16 May 2025. The Summer School provided a platform for PhD candidates to engage in dynamic discussions and showcase their research. It took place at Tübingen castle and featured expert lectures from leading scholars and practitioners, participant presentations, and social events in view of fostering in-depth discussions amongst participants.

The Summer School was stacked with a stellar line-up of speakers, including Mateja Durović (ECHR and King‘s College London), Melanie Fink (Leiden University), Marta Cantero Gamito (University of Tartu), Matthias C. Kettemann (University of Innsbruck), Mariana Valente (University of St. Gallen), Kyoko Yoshinaga (Keio University).

Michèle Finck recently had the opportunity for a direct discussion with the German President Frank-Walter Steinmeier. Their discussion covered critical topics such as AI regulation and its effective communication, the importance of interdisciplinary research, and strategies for building and fostering trustworthy AI.

This discussion occurred as part of the event series Werkstatt des Wandels hosted by Federal President Frank-Walter Steinmeier. 

The event featured other prominent participants, including Michael Bolle, Chairman of the Foundation Board of the Carl-Zeiss-Stiftung; Gisela Lanza, Professor and Director at the WBK Institute of Production Science at KIT Karlsruhe; Boris Otto, Director of the Fraunhofer Institute for Software and Systems Technology (ISST) and Professor of Industrial Information Management at TU Dortmund; Philippe Souidi, SAP Vice President; Dr. Michael Fausten, Senior Vice President of Corporate Research and Division Head for Artificial Intelligence and Systems at Bosch; Vanessa Cann, Managing Director at Accenture; and Stefan Hartung, CEO at Bosch.