DigiPub 2026: connecting young researchers
DigiPub 2026 took place from 8. to 10. Juli 2026 at the University of Basel: an international in-person academic workshop focusing on the digital transformation of public administration.
This year, the emphasis was on legal methodology in public law. It brougth together early-career researchers from Europe and beyond for a series of presentations, group discussions, and collaborative sessions.
Participants did present their ongoing research, received targeted feedback, and engaged with key challenges in digital governance. The program was complemented by keynote lectures by leading scholars.
DigiPub 2026 on LinkedIn
(in order of appearance)
| Post of 16. July 2026: 01 presenting Ljubiša Metikoš | At DigiPubCon 2026: Ljubiša Metikoš analyses different approaches to the explainability of judicial AI. Although various frameworks - such as due process standards, data protection regulations, and AI governance frameworks - establish a foundation for the right to an explanation regarding judicial AI, significant interpretative gaps remain. These gaps can be addressed through theories of procedural justice before judicial AI can be fully deployed. Post and picture by Sebastian Rous |
| Post of 17. July 2026: 02 presenting Magdalena Eder | At DigiPubCon 2026: What is the meaning behind «risk», which risks must be addressed, and to what level should risk be mitigated under the AI Act? Magdalena Eder’s work engages precisely with these questions. The concept of «risk» under Article 9 of the AI Act, borrowed from the field of product safety regulations, can be understood as a function with two variables: the probability that harm will occur, and the severity of harm if it does occur. The problem is that the harm addressed by this article consists essentially of fundamental rights infringements, and this is where the product-safety framework that the concept is based on becomes insufficient. Magdalena thus argues that we should develop a more rigorous and context-specific operationalization of terms within such a rapidly evolving field. Post by Sabina-Andreea Grigore, picture by Apollo Dauag |
| Post of 20. July 2026: 03 presenting Maria Broini | At DigiPubCon 2026: «When an algorithm makes a decision that causes harm, who's actually responsible?» Carolina Broini Marcondes tackled this question head-on in her presentation «Artificial Intelligence: When the Creation Crosses the Line.» She identified a central feature of our justice system as being designed to judge individuals who are capable of intention, understanding, and guilt. However, machines lack the ability to feel guilt or comprehend punishment. Therefore, when AI becomes part of misconduct, the law begins to reach its limits. One easy answer is to blame the manufacturer. But Marcondes pushes back on that not because it is wrong, but because it is incomplete. Blaming the manufacturer fails to address the fundamental issue that our entire framework of responsibility was never constructed for a world in which the «actor» is not human at all. Drawing upon the EU AI Act, she asserts that criminal law, artificial intelligence, and public administration are now intersecting in unforeseen ways. She emphasizes the necessity to reevaluate the concept of accountability entirely, rather than merely applying superficial fixes. Post by Ifeanyi Ogbodo, ABR., picture by Vincent Mesmer |
| Post of 22. July 2026: 04 presenting Clément Albaret | At DigiPubCon 2026: Public administrations increasingly rely on «public algorithms». Clément Albaret examines, how existing legal review mechanisms (especially judicial review) are partly ill-suited for public AI-systems. He argues that algorithms reshape legal concepts and administrative law. This creates a two-way relationship: law governs algorithms, while automation transforms legality itself. The research fills a gap by examining these feedback effects on the conceptual structure and practice of public law. Post and picture by Lonnie Petersen |
| Post of 24. July 2026: 05 presenting Ifeanyi Ogbodo | At DigiPubCon 2026: In his presentation, Ifeanyi Ogbodo, ABR. examines the practical implementation and effectiveness of the Nagoya Protocol in governing access to genetic resources and benefit-sharing, alongside newer mechanisms like the Cali Fund for the Fair and Equitable Sharing of Benefits from the use of Digital Sequence Information on Genetic Resources and the WIPO Treaty on Intellectual Property, Genetic Resources and Associated Traditional Knowledge. He identifies significant gaps, including weak enforcement, technological evasion, and misaligned incentives among stakeholders. While the Protocol embodies noble ideals of equity and conservation, its implementation faces obstacles that undermine its intended impact. Post and picture by Yquem Zberg |
| Post of 27. July 2026: 06 presenting Emanuele Brasca | At DigiPubCon 2026: Emanuele Brasca shines a light on how observing the application of AI through «the lenses of law» can help to identify significant obstacles to the fruitful implementation of AI in public procurement and highlights the importance of an European exchange of best practices and the creation of collaborative administrative procedures through digitalization. Post and picture by Tobias Wöhner |
| Post of 29. July 2026: 07 presenting Karol Wundzinski | At DigiPubCon 2026: Smart Hybrid Method! Karol Wundziński presented his work on the algorithmic case flow management and case weighting in the Polish administrative judiciary. As many other courts as well, they face the problem of backlog which makes efficient case allocation a core issue. Today, cases are still allocated manually with the assumption that every case take the same amount of work – which is far from reality. Karol looks at the methods that other jurisdictions use to assign cases efficiently. This includes the time-study method in Germany and Austria, the Delphi Method in Demark, and a hybrid method in the Netherlands. Drawing on inspiration from these other methods, he is developing a new one: the «smart hybrid method». This method will combine empirical data with expert knowledge by integrating automated AI-powered analysis with the Delphi method. Apart from this interesting approach, I was also impressed by how aware Karol is of the implementation reality of his approach and how much he takes care of designing something that is useful for the judges. Post and picture by Selina Scherrer. |
| Post of 31. July 2026: 08 presenting François Péan | At DigiPubCon 2026: François Péan from the University of Angers examines «algorithmic discrimination» in public law, beginning by pinning down what the term actually means. He shows how central the notion of indirect discrimination is to defining it, and explains why scaling and personalisation are specific to algorithmic discrimination – and what legal challenges follow: particularly the difficulties for effective regulation and litigation, and the question of how to measure and correct algorithmic outcomes. I am looking forward to reading more of his work on this topic in the future. Post and picture by Evelyne Putz |
| Post of 3. August 2026: 09 Post Evelyne Putz | At DigiPubCon 2026: We were delighted to hear the presentation of Evelyne Putz from the University of Graz. She explained, with a case view analysis focus on the European Digital Identity Wallet, the methodological challenges of implementation of technology-neutral requirement. Her presentation arises the fact that technology-neutral regulation requires not just a technical execution but that the legal analysis must follow the norm into the system. Answering that issues require traceability of the implementation of the norm on the technical system. We are looking forward to reading are very interesting work. Post and picture by François Péan |
| Post of 5. August 2026: 10 Post Selina Scherrer | At DigiPubCon 2026: Selina Scherrer is conducting research on how citizens seek legal information online. In her project, she will combine a dogmatic approach with an empirical study to gather information on people's real-life behaviour and the legal conditions governing access to legal information. She claims that supplementing traditional methods of legal research with empirical methods – for example, those drawn from the social sciences – can elevate the discipline and broaden its scope. Post and picture by Karol Wundziński |
| Post of 7. August 2026: 11 Post Tobias Wöhner | At DigiPubCon 2026: Tobias Wöhner explains the actual role of data protection and why European digital regulation needs more legal dogmatics. Since data protection law became more and more central in the european context, many tensions, questions and challenges have arised! What I bring in my knowledge luggage: the need to rethink the role of data protection law, promoting the legal dogmatic way of thinking also in this digital age! Post and picture by Emanuele Brasca |
| Post of 10. August 2026: 12 Post Lonnie Petersen | At DigiPubCon 2026: You have just received a decision made automatically by your government agency, but you don't understand the reasons behind it. What can you do? Lonnie Petersen interrogates transparency conformity through the lenses of its three components : traceability, explicability, and communication. Her presentation showed that although the rules are technologically neutral and formally grand access to information, their underlying assumptions are challenged by digital decision-making systems. Post and picture by Lucile Pasche |
| Post of 12. August 2026: 13 Post Sabina Grigore | At DigiPubCon 2026: Sabina-Andreea Grigore shared valuable insights into the evolving role and responsibilities of prosecutors investigating atrocities in present times. Her research focuses on newly established technological mechanisms, particularly in situations where European institutions are not always open about how these technologies operate and why double structures exist. By combining comparative, empirical, and legal research methods, she seeks to fill this gap and shed light on these otherwise opaque processes. Post and picture by Magdalena Eder |
| Post of 15. August 2026: 14 Post Sebastian Rous | At DigiPubCon 2026: Sebastian Rous provides a taxonomy of the concepts of synthetic content, media and data across the EU Digital Acquis, including the GDPR, DSA, the AI Act, and the Data Act. These can refer to AI-generated media, training data for AI systems, audio, video, text, etc. all depending on the specific legal context. Sometimes criteria such as the 'deceptive capacity' is used. Moreover, the different concepts mentioned across these laws also overlap in their scope, but then again with particular exceptions. To provide more clarity in this regard, Rous' taxonomy aims to structure and classify these concepts. In the discussion with the audience Rous also talked about the political negotiations during the trilogues that lead to the development of these concepts in the AI Act. Post and picture by Ljubiša Metikoš |
| Post of 17. August 2026: 15 Post Lucile Pasche | At DigiPubCon 2026: Lucile Pasche (🔎 data protection law) discovered a practical challenge for private companies entrusted to carry out public tasks on behalf of the Confederation (Art. 5 lit. i FADP). How to ensure compliance with the standards rising from federal public AND private law? Lucile presented some solutions, depending on the size of the company: from one person with multiple responsabilities to multiple persons for the two tasks. In the end it's more a matter of practice, not the law: «We do not need better data protection law, but a better practice.» Regarding GDPR (see picture) she said «being in Switzerland I could not _not_ look at it». Post and picture by Apollo Dauag |
| Post of 19. August 2026: 16 Post Yquem Zberg | At DigiPubCon 2026: «A child is drowning in a pond right in front of you. Is there something wrong with asking ChatGPT how to act?» This is the question Yquem Zberg sets out to answer in his research in legal philosophy on delegating moral judgment to LLMs. Drawing on Kantian moral philosophy and philosophy of mind literature such as Searle's Chinese Room thought experiment, he argues that LLMs lack moral autonomy, a necessary condition for exercising practical reason and therefore for moral judgment itself. Such a view implies a «duty of moral authenticity»: a duty to exercise our own practical reason. This duty, he writes, «is violated when we consult LLMs on moral questions as if they had practical reason». If asking ChatGPT moral questions is a failure to take our own autonomy seriously, Yquem doesn't conclude that AI has no legitimate role in moral life. A tool can support practical reasoning, but it should never replace our authentic judgment as moral agents. Post and picture by Clément Albaret |