EU digital law is becoming increasingly complex. The AI Act, with its intricate web of responsibilities, oversight, and enforcement at the EU and national levels, illustrates this very clearly. The interdisciplinary paper by Magdalena Stratmann, Dr. Tobias Mast, et al., which emerged from the project “Informing Regulatory Reasoning on Algorithmic Systems in Societal Communication with STEAM” (STEAM), introduces Architectural Ecosystem Modeling—a visual method designed to reveal the institutional structures of the AI Act, analyze responsibilities and dependencies, and identify gaps, overlaps, and opportunities for coordination within the regulatory framework.
Abstract
European Union (EU) digital law is increasingly characterized by complex multi-level regulatory structures that challenge traditional legal analysis. The Artificial Intelligence (AI) Act exemplifies this trend, with its broad range of substantive provisions that envisage diverse mechanisms of oversight, implementation, and enforcement by different– often interdependent – institutions at both the EU and national levels. This article introduces architectural ecosystem modeling as a novel, visualization-based method for mapping and analyzing the institutional regulatory structures embedded in EU legal acts. Focusing on the AI Act, we demonstrate how this approach can reveal the distribution of regulatory roles, responsibilities, and interdependencies among different actors, enabling clearer insights into the Act’s implementation and oversight framework. In distinction to other approaches, our approach does not aim to support individual compliance assessment but instead targets, in particular, EU and national authorities, policymakers, and legislators seeking to understand the regulatory structures of the AI Act. By visually representing complex regulatory arrangements, architectural ecosystem modeling can help identify regulatory redundancies, gaps, and opportunities for improved cooperation and coordination. We further explore how the approach could be applied more broadly, how it might be enhanced by digitalization and how it might ultimately be scaled through automation.
Stratmann, M., Mast, T., Spürkel, J., Burmeister, F., Kurtz, C. Soulier, E.: Visualizing Regulatory Ecosystems: A New Approach to Mapping EU Regulations as Architectures – The Case of the AI Act. In: Computer Law & Security Review (61/2026). https://doi.org/10.1016/j.clsr.2026.106333