Berkes RudolfOn 29 May 2025, the Algorithmic Constitutionalism Lendület/Momentum Research Group gathered for its inaugural in-house conference to discuss the seven manuscripts currently under preparation as part of the project. These working papers reflect on the impact of artificial intelligence on three fundamental rights - right to fair trial, freedom of speech, and right to a healthy environment. The working papers were reviewed by invited expert colleagues who presented their recommendations during the discussions of the early results. The present report aims to summarize the different strands of academic endeavours our Research Group is currently pursuing.
The first presented and reviewed manuscript focused on the responsible Use Of AI In Parliamentary Settings – With A Particular Weight On Automated Speech Recognition. This manuscript has been co-authored by principal investigator Boldizsár Szentgáli-Tóth, by Rudolf Berkes, researcher working full time on the project, and by Klára Szalay, researcher of both the Institute for Legal Studies and the Office of the Hungarian Parliament.
This working paper examines how parliaments are implementing AI technologies, with a particular focus on automated speech recognition (ASR) systems. The authors conducted a comprehensive review of relevant parliamentary innovations reported by the Inter-Parliamentary Union (IPU) and other international organizations, supplemented by academic literature on ASR technology and AI governance. The analysis applies an ethical framework derived from international AI governance principles to assess current implementations and identify best practices. The paper was reviewed by Miklós Sebők, Research Professor of the HUN-REN Institute for Political Science.
The discussion continued with the review of a paper dealing with Generative artificial intelligence in the judiciary: Some arguments for a responsibility-based interpretative approach. This manuscript has been co-authored by principal investigator Boldizsár Szentgáli-Tóth, by Dóra Pálfi, doctoral student at Károli Gáspár University of the Reformed Church, and by Kinga Kálmán, doctoral student at Eötvös Loránd University.
This study analyses the diverse approaches to regulating GenAI use in courts across various jurisdictions, examines their underlying principles and practical implications, and proposes an interpretative approach that balances innovation with the fundamental values of the legal system. A multi-faceted research methodology has been employed for analysis from the perspectives of different actors involved in judicial proceedings.
The research methodology combines comprehensive case law analysis with systematic examination of the evolving landscape of guidelines, standing orders, and ethical frameworks related to GenAI in legal practice across multiple jurisdictions. The authors aimed at conducting a global overview of regulatory approaches, involving detailed analysis of official guidelines from various authoritative bodies, including the Council of Europe (CEPEJ), the United Kingdom judiciary, New Zealand and Caribbean legal systems, and numerous United States courts. Through this comparative legal analysis approach and the development of a typology of regulatory responses, the paper examines how different stakeholders—courts, lawyers, and unrepresented parties—interact with GenAI in judicial contexts. The paper was reviewed by Péter Homoki, lawyer, and regulatory specialist.
The third discussed manuscript covered the link between Artificial Intelligence and the Right to a Healthy Environment. This manuscript has been co-authored by Gábor Kecskés, research fellow at the Institute for Legal Studies, and by Ágnes Tahyné Kovács, associate professor at the Pázmány Péter Catholic University.
This working paper argues that the foundational principles of international environmental law can serve as an analogy for developing legal frameworks for artificial intelligence (AI) applications. This premise stems from the observation that both fields, despite their relatively short legal histories, are rapidly evolving due to technological advancements, necessitating swift legal responses. Moreover, both domains are complex and multifaceted, posing diverse regulatory challenges. The study outlines the evolution and characteristics of the right to a healthy environment and highlights the environmental implications of AI applications. It then explores key environmental law principles and their relationship with the Sustainable Development Goals (SDGs) and AI applications. After addressing the tensions that remain in applying this analogy, the study concludes with recommendations based on the research findings. The paper was reviewed by András Pünkösty, assistant professor of the Pázmány Péter Catholic University, and research fellow of the Ludovika University of Public Service.
Going further with the environmental aspect, the next academic contribution analysed the potential regulation of AI-based Smart Metering in the EU and Three Member States with a Legislative Quality Assessment. This manuscript has been co-authored by Zsófia Biro, research fellow at the University of Pécs, by Dorottya Deáki, research assistant at the Institute of Legal Studies, and by Luka Martin Tomazic, associate professor at Alma Mater University of Maribor.
This working paper aims to uncover the regulatory surroundings of smart meters within select Central and Southeast European countries, exploring some particularities and underlining good practices and legislative challenges while considering the relevant EU framework. Through comparative analysis, the authors aim to evaluate and identify the most optimal approach regarding smart metering that captures the ideal relationship and balance concerning fundamental rights. In achieving that, they explore the relevant case law and utilize testing tools to determine the quality of legislation used. The study exclusively examines the EU regulatory framework and the Hungarian, Slovenian and Croatian regulations on this subject, focusing solely on residential consumers and the application of smart meters in the context of electricity. The paper was reviewed by Lilla Király, associate professor at the Károli Gáspár University of the Reformed Church.
In the field of freedom of speech, the fifth paper had the following title: artificial intelligence and human contribution in content moderation: revisiting the differentiation between factual statements, value judgments and narratives. This manuscript was resulted by the collaboration of four experts: principal investigator Boldizsár Szentgáli-Tóth, by Gábor Polyák, senior research fellow at the Institute for Legal Studies, by Rudolf Berkes, researcher working full time on the project, and by Orsolya Ferencz, doctoral student of Andrássy University.
This working paper aims to reconsider certain aspects of the collaboration between artificial intelligence (AI)-based tools and human agents conducting content moderation in virtual platforms. According to the authors’ hypothesis, the distinction between misinformation, malinformation and disinformation should be read in conjunction with the differentiation between factual statements, value judgments and narratives. Then, based on this complex classification, paramount conclusions might be drawn about how to exploit AI in the most effective way filtering illegally conveyed content from virtual platforms. The paper was reviewed by Péter Bányász, associate professor at the Ludovika University of Public Service.
Dániel Eszteri, Head of Department in the National Authority for Data Protection and Freedom of Information, and Péter Sziládi, lawyer specialized in data protection and labour law, conceptualized the impact of artificial intelligence on the protection of personal data in the field of employment.
This manuscript investigates the intersection between AI technologies and employee data protection obligations under the GDPR. The applicable legal norms, the evolving interpretation of key GDPR principles, and the regulatory responses at both EU and national levels were reviewed. To illustrate the complexity of the problem, the national legal frameworks of three EU Member States: Germany, Italy, and Hungary were also briefly examined. These countries were selected because they represent slightly different civil law traditions and regulatory cultures within the EU: Germany with its particularly strong constitutional emphasis on human dignity and personality rights as well as strong collective labour law, Italy with its long-standing restrictions on employee surveillance rooted in labour law traditions, and Hungary with a more employer-favourable but still GDPR-aligned framework. This comparative angle allows to highlight both common trends and notable divergences in regulating AI-based workplace monitoring. This manuscript was reviewed by Gergely Lendvai, research fellow at the Center for Advanced Internet Studies, Bochum.
The last discussed academic piece was again prepared by four co-authors: principal investigator Boldizsár Szentgáli-Tóth, Kitti Mezei, senior research fellow at the Institute for Legal Studies, Anikó Träger, researcher at the same institute, and Rudolf Berkes, researchers at the Institute for Legal Studies. This manuscript analyzed the Legal implications of TikTok’s influence during the most recent elections around the world, with special regard to the canceled 2024. Romanian presidential elections.
This working paper aims to contribute to the literature by assessing whether the management of TikTok has complied with its obligations under the EU’s Digital Services Act to protect the integrity of democratic elections during campaign periods. The paper highlights the major reasons that caused the political exploitation of TikTok in Romania and by formulating some policy recommendations based on this assessment. According to the authors’ hypothesis, among other reasons, the lack of effective sanctions on platform operators by the DSA and the weaknesses of TikTok’s content moderation mechanisms led to the massive undue spread of political advertisements for the clear benefit of a presidential candidate during the campaign period. The paper was reviewed by Gergely Gosztonyi, associate professor of the Eötvös Loránd University.
Each project contributor is grateful for the reviewers’ and audience participants’ comments during the discussions, they provided many invaluable insights for the manuscripts. The conference also provided a new arena for interdisciplinary discourse on artificial intelligence’s diverse impact on human rights. The next step is the finalization of the manuscripts based on the received comments and suggestions, preparing them to be submitted high-ranking journals (Q1-Q2) in Scopus by the end of summer 2025.
__________________________________________________________
This report was prepared with the support of the Algorithmic Constitutionalism Research Group (LP2024-20/2024), funded by the Hungarian Academy of Sciences.
__________________________________________________________
The views expressed above belong to the author and do not necessarily represent the views of the Centre for Social Sciences.