5th German Conference for Young Researchers in Private International Law in Heidelberg – Conference Report
Written by Victoria Hélène Dintelmann (Heidelberg University)
On February 14th and 15th, 2025, more than one hundred young academics gathered at Heidelberg University for the 5th German Conference for Young Researchers in Private International Law to discuss the topic “Digital Transformation and Private International Law – Local Connections in Boundless Spaces”. The conference was organized by Andreas Engel, Sophia Schwemmer, Felix Berner, Aron Johanson, Markus Lieberknecht, Ann-Kathrin Voß, Charlotte Wendland and Anton Zimmermann.
The first day started with Professor Marc-Philippe Weller (Heidelberg University), director of the Institute for Comparative Law, Conflict of Laws and International Business Law, illustrating Heidelberg University’s Private International Law tradition. For instance, Max Gutzwiller, who rejected renvoi as well as party autonomy in Private International Law, was the director of the Institute from 1929 until he was forced to emigrate to Switzerland in 1935. Weller ended his remarks with special emphasis on the late Erik Jayme, whose impact on Private International Law was vast. For example, Jayme advanced the “two-stage theory of Private International Law”. Further, he introduced postmodern thoughts of mobility, multiculturalism and openness to Private International Law, arguing for every human to have a “droit à la difference”.
Professor Christiane Wendehorst (University of Vienna) gave the keynote lecture on digital goods in Private International Law. She focused on the Private International Law treatment of digital goods regarding rights with third-party effects. In her introduction, she differentiated between digital goods based on their level of exclusivity and the ability to duplicate them. Within crypto assets in particular, Wendehorst differentiated between tokens with an internal value such as bitcoin (“intrinsic tokens”) and tokens that represent an asset outside the crypto system (“extrinsic tokens”). She deemed this differentiation to be of great importance to assess the applicable law: for extrinsic tokens, the statute of the represented asset must be considered. While some tokens are regulated, e.g. by Sec. 32 of the German Electronic Securities Act, Wendehorst expressed criticism towards an analogous application of such provisions, doubting the tokens’ functional comparability. She then continued with a comparative approach and illustrated different national laws as well as international attempts at a more uniform Private International Law approach to rights in rem to digital assets. She emphasized rules under which a choice of law regarding rights with third-party effects is possible. For instance, the rules of the United States’ UCC refer to the lex fori of the District of Columbia in absence of a choice of law as a fallback. A similar approach, looking first at a choice of law and last at the law of the forum state, was adopted under Principle 5 of the UNIDROIT Principles on Digital Assets and Private Law. Wendehorst concluded by explaining the purposes of the different approaches. In the end, Wendehorst made the plea for a more comprehensive solution and ideally more uniform conflict of laws rules to solve what she called a “crisis in International Property Law”.
Johannes Weigl (LMU Munich) presented on data-related European conflict of laws questions. He first showed that the decades-old “libertarian dream” of a boundless internet did not come to fruition: data is regulated by states. Still, digital and analogous goods cannot be equated, leading to a call for a harmonized digital property law. Such a uniform law would cause the “silent death” of conflict of laws provisions regarding digital property. Still, Weigl identified four categories in which questions of conflict of laws might nonetheless arise. As to territorial limits of harmonization, he identified as a first category the territorial scope of EU digital regulation and as a second category data protection through the limitation of the free flow of data beyond the EU’s borders. Regarding the substantive limits of harmonization, he considered a third category of potential conflict of laws challenges to be explicit references to national law and, as a fourth, substantive gaps of uniform law. Weigl went on to discuss limits of boundlessness using the examples of his first and third category. Regarding the territorial scope of EU digital regulations, many do not depend on the provider’s place of establishment but on whether the services are offered to persons in the EU. While Weigl classified those as one-sided conflict norms undoubtedly belonging to public law, he argued for their parallel application as public and private law conflicts rules. Weigl explained this approach to be – above all – teleologically convincing, securing the effet utile of EU law as well as international decisional harmony between public and private law. Further, Weigl illustrated the substantive limits of unification using the example of the third category, i.e. rules explicitly referring to national law. While some see such referential norms as conflict of laws rules, he argued against this classification, maintaining that referential norms are not conflict of laws rules but leave room for general conflicts rules. As this approach leads to the application of general conflict of laws rules, he identified some room for a more general legal policy discussion, e.g. about further harmonization of conflict of laws rules or the creation of internet specific conflicts rules.
Loïc Bréhin (Université Panthéon-Assas) addressed the law applicable to determine the illegality of digital content. Pursuant to Art. 3(h) DSA, content is illegal if it is not in compliance with EU law or the law of Member States. Bréhin criticized this provision as too generic; it does not determine the applicable law. He identified the root of the problem to be the diversity of legal relationships one could assess: there is a relationship between victim and publisher, victim and platform, as well as publisher and platform. Bréhin explained that to all relationships, different rules may apply and thereby cause inconsistencies. Bréhin acknowledged that the problem could be mitigated by solutions at the edge of conflict of laws theory such as internal market clauses or through fundamental rights. However, he found the most promising solution to lie at the heart of conflict of laws theory: substantive law consideration. He proposed to assess the legality of content under the law designated by the conflicts rule for torts invokable by the victim, either as applicable law or as law to be taken into consideration at the level of substantive law. Bréhin based this proposal on the rationale of Art. 3(h) DSA and Art. 14(4) DSA, maintaining that although digital platforms are often classified as private, they are in fact collective phenomena. He concluded that there is great potential in allowing for adjustments – in particular, when considering the platform’s nature as a collective phenomenon.
Christina Lemke (University of Hamburg, Max Planck Institute for Comparative and International Private Law Hamburg) tackled questions regarding the implementation of the digital euro as a European digital currency from a Private International Law perspective. Lemke introduced the topic by differentiating between cash, electronic money and the digital euro. She classified cash, on the one hand, to be a central bank liability to which individuals have property rights. Electronic money, on the other hand, is a means of payment that derives its value from a claim against a private institution. Lemke explained that in contrast, the digital euro is to be a central bank liability, aimed at supplementing cash payment. Neither the technological details nor the digital euro’s legal nature are certain. Lemke maintained that the digital euro should not be classified as a mere claim, since it can be allocated to an individual. Lemke determined the most important question in relation to the digital euro to be the function of payment, i.e. the evaluation of the satisfaction of payment obligations. The first step in answering this question is the determination of the applicable law. To assess payment, one could look at the lex causae. Lemke emphasized the importance of the lex monetae principle for monetary units: Anchored in sovereignty, every state is entitled to its own currency. Hence, a monetary unit is governed by the sovereign that issued the unit. However, the digital euro is not a monetary unit, but a monetary medium. Lemke argued for the extension of the lex monetae principle to the monetary medium. Lemke concluded by raising the delicate questions on the EU’s competence to develop private law regulations on the digital euro and the conflicts between EU institutions possibly involved.
Naivi Chikoc Barreda (University of Ottawa) elaborated on the rise of remote authentic instruments when notarizing beyond borders through online appearance. While notarial practice is increasingly shaped by digitization, there is potential for conflict when a party is in a different country than the notary. Chikoc Barreda started by giving a comparative overview of the three main approaches to deal with remote authentication: first the liberal approach, which allows all relations to be handled remotely, second the intermediate approach, which allows for exceptions in very protected fields of law (e.g. wills, divorces) and third the restrictive approach, which generally prohibits remote authentication with few exceptions (e.g. the incorporation of companies). Chikoc Barreda explained that this fragmentation leads to challenges for Private International Law. One of these challenges is to assess whether the locus actus is the state where the notary is located or the state from which the parties appear. While jurisdictions following the liberal approach view the location of the notary as decisive, restrictive jurisdictions tend to prioritize the state from which the parties appear. This leads to the risk of limping legal relationships. Further, Chikoc Barreda showed that questions of equivalence of acts arise. Authenticity relies on a person’s assessment by the notary. The classic notion was to reach such an assessment through physical presence. Under a more modern approach, in some jurisdictions, virtual presence suffices. In light of this, Chikoc Barreda elaborated on the assessment of the equivalence of notarial acts: while the state of origin will regularly apply the lex auctoris to determine equivalence, the receiving state might apply another law to the form. Last, Chikoc Barreda addressed the notary’s international competence: some view a foreign notary as having unrestricted competence in line with the principle of free choice, while others only accept a restricted competence of the notary, demanding for a significant connection to the notary’s state of origin. Chikoc Barreda concluded that the rise of remote authentication calls into question the lex loci actus rule, authenticity, and the notary’s international competence.
Piotr Wilinski (Erasmus University Rotterdam) and Marciej Durbas (KKG Legal, Kraków) discussed the consequences of the use of AI by arbitral tribunals – in particular, potential challenges of arbitrators and awards. Wilinski and Durbas first introduced the legal framework, stating that there is no significant transnational law governing the use of AI in arbitration. However, there are emerging legal instruments, e.g. in the EU and the US. The EU AI Act governs individuals who rely on AI as deployers. A deployer status causes a duty to disclose. Wilinski and Durbas argued that arbitrators can be classified as deployers within the meaning of the EU AI Act, causing potential disclosure obligations. At the same time, there is only nascent soft law, namely the Silicon Valley AI guidelines and the SCC guidelines. These rules are quite rudimental. Wilinski and Durbas agreed that under the guidelines, decision-making may not be delegated to AI. Second, Wilinski and Durbas turned to potential challenges of arbitrators. They found that AI can be used to assist decision-making. Although most tasks one might delegate to AI do not directly affect decision-making, it does seem possible that steps such as AI-generated summaries of cases indirectly affect the decision. Wilinski and Durbas proposed that an improper use of AI could lead to challenges of the tribunal. Third, Wilinski and Durbas assessed the enforceability of awards rendered with the use of AI. Although AI is a new phenomenon, Wilinski and Durbas argued that the core of the problem is not. They drew a comparison of the use of AI on the one hand with the use of tribunal secretaries and independent legal research by arbitrators on the other hand. Based on this comparison, they deduced that as long as AI is merely used for assistance with the award’s drafting (even if its use was undisclosed), the award will likely stand. When it comes to decision-making, AI may be used for support in reasoning, but they found that to secure enforcement, the decision itself must stay with the tribunal. Wilinski and Durbas concluded that for now, as long as AI does not render the final decision, arbitrators can “sleep safely”. However, they found a common standard to be preferrable, perhaps in the form of a traffic light approach.
The last speaker of the first day was Agatha Brandão (University of Luzern), who presented on the development of a large language model for Swiss cases on choice of law (available at https://www.choiceoflawdataverse.com). The project’s goal was to use an open AI GPT to generate high-quality case law analysis comparable to Private International Law experts. Using a data set of 33 cases, the AI was to perform six tasks: to extract an abstract, to extract and summarize relevant facts, to extract the relevant Private International Law provisions, to classify and interpret the choice of law issue and to extract and interpret the court’s position. Brandão maintained that the AI case analyzer succeeded in the extraction and classification of information. However, challenges arose when the AI case analyzer provided information that was secondary or irrelevant and when it produced lengthy responses. Brandão explained that in working on fixing these problems, the research team focused on phrasing prompts as precisely as possible: if the output did not match the researchers’ expectations, the instructions were most likely not sufficiently comprehensive. At the end of the experiment, each category of tasks was evaluated based on specific criteria in a peer-reviewed process. Overall, the AI case analyzer had a success rate of 92 %. While there were still roughly 10 % of outcomes one might want to modify, Brandão emphasized that the AI case analyzer saves valuable time – in particular, for the extraction and classification of information and when given sufficiently precise instructions. Brandão concluded that large language models can indeed be a valuable support – not unlike real-life Private International Law experts.
The second day of the conference started with parallel panel discussions. In the first panel, Christoph König (BSP Berlin) gave an impulse rooted in legal history on the decentralization of blockchain technology and delegalization. König drew parallels from discussions surrounding the creation of a lex mercatoria in the past century. The second panel focused on the pioneering role of arbitration in the use of digital tools in contrast to the use of digital means in German and Swiss courts. First, Cedric Schad (University of St. Gallen) gave an overview over the advanced, but not boundless use of digital instruments in arbitration. In particular, he illustrated the option of conducting proceedings via video conference and the use of case management platforms. Second, Marco Andjic (Osnabrück University) presented on attempts at digitization in German courts: he found that the main obstacle of remote proceedings is not German law, but the equipment of courts. Third, Nadine Boss (University of St. Gallen) elaborated on the Swiss approach. While there is no option of virtual court proceedings yet, there are attempts at reform. It is possible to use digital tools such as e-mail, but uncommon due to perceived risks regarding service. In the third panel, Raffael Müller (Heidelberg University) presented on international product liability and AI. Müller considered the applicability of Art. 5 of the Rome II Regulation to Artificial Intelligence. He emphasized the importance of placing AI on the market and its interplay with the AI Act, in particular regarding the AI Act’s territorial scope. Fourth, Peter Moser (LMU Munich) addressed connecting factors for declarations of intent made by AI. Moser differentiated between an “ePerson” and an “AI agent”. An “ePerson”, on the one hand, can be legally competent and capable. As Art. 7 of the Introductory Act to the German Civil Code concerns natural persons, Moser found that a corporate law connecting factor might be more appropriate. An “AI agent”, on the other hand, is no proper legal entity. Hence, the attribution of its actions is critical. Moser found it most appropriate to apply Art. 10 Rome I Regulation, as the exclusion in Art. 1(2)(g) Rome I Regulation concerns natural persons – not an “AI agent”. In the fifth panel, Leon Marcel Kahl (University of Vienna) illustrated how the special construction of the Unified Patent Court leads to conflict of laws questions. Which conflict of laws rules the Unified Patent Court applies is determined by a “ladder” in Art. 24(2) UPCA. According to its lit. c, the applicable national conflicts rules are to be determined by the court. However, since the Court of First Instance comprises a central chamber as well as local and regional chambers, it is not clear which national provisions are to be applied.
After the panel discussions, Linda Kuschel (Bucerius Law School) elaborated on whether cross-border electronic service is a sovereign act on foreign territory. In Germany, regular e-mails do not suffice for proper service, but the use of a special electronic attorney mailbox (“beA”) does. Internationally, there are cases of service through e-mail and even social media platforms. First, Kuschel identified the European Service Regulation and the Hague Service Convention as the relevant rules for cross-border service. Next, she discussed the Public International Law qualification of service. The prevailing opinion considers the service of court documents to be an exercise of state authority. This is where Kuschel differentiated: while she qualified the legal consequences of service as an exercise of state authority, she did not find the same to apply to the mere act of gaining knowledge of a document and its content, e.g. through service by private means. She then tackled the question of localization of electronic service. First, one could see electronic service as a type of fictional service. But while fictional service is a mere last resort, electronic service could become the norm – therefore, Kuschel negated a comparability. Second, one could view the internet as an exterritorial space that cannot be attributed to any sovereign state, but the internet is not truly boundless. Third, one could draw an analogy to analogous life and treat electronic service parallel to analogous service, as territorial borders are emulated in the digital space. However, equating analogous and electronic service would lead to a fiction. Kuschel assessed this to be particularly problematic if one – in line with the prevailing opinion – classifies service in a foreign state as an act of sovereignty on foreign territory. In light of these shortcomings, Kuschel deemed it necessary to assess electronic service by its own metrics. She concluded that only service on foreign territory through means of sovereign power leads to a violation of the principle of territoriality while in contrast, service by means of communication accessible to private persons should not violate Public International Law.
The last presentation was delivered by Adrian Hemler (University of Konstanz), who illustrated options and boundaries of a fully digital judicial activity from abroad. Hemler reported a trend towards virtual and digital proceedings, asserting that these developments can only be expected to accelerate. The advantages in virtual proceedings lie in more efficiency, lower costs as well as in making the profession of judge more flexible and, hence, more attractive. While Hemler found possibly affected principles of German procedural law to be publicity, immediacy and orality, he assessed that their violation can be avoided. However, Hemler explained the currently prevailing opinion to be that working from abroad as a judge violates the foreign country’s sovereignty. Hemler went on to reference Kelsen, who understood what ought to be at the core of law – not what is. Building on this, Hemler differentiated between on the one hand the scope of application of legal norms, which operates on the level of what ought to be. This category does not violate foreign sovereignty, even when it extends beyond a state’s territorial borders. On the other hand, Hemler allocated the practical implementation and enforcement of legal rules on the level of what is. Hemler argued that this latter category should only be allowed with the other country’s permission – otherwise, Public International Law violations can arise. Within this grid, according to Hemler, rendering judgements from abroad does not interfere in the foreign state’s sovereignty.
A conference volume will be published by Mohr Siebeck later this year. The 6th German Conference for Young Researchers in Private International Law will take place at LMU Munich in 2027.