University of Tehran on behalf of the "Cyberspace Research Policy Center" and the "UNESCO Chair on Cyberspace and Culture: Dual Spacization of the World" Journal of Cyberspace Studies 2588-5499 10 2 2026 07 01 From Pretoria to Brussels: Expert networks, ethical spillover, and the architecture of EU AI regulation 449 472 105007 10.22059/jcss.2025.403975.1187 EN Ali Sabbaghian Department of Regional Studies, Faculty of Law and Political Science, University of Tehran, Tehran, Iran. Rajdeep Singh Department of Regional Studies, Faculty of Law and Political Science, University of Tehran, Tehran, Iran. 0000-0003-1218-134X Journal Article 2025 10 09 <strong>Background:</strong> As artificial intelligence (AI) technologies become embedded in public and private decision-making, questions of governance have become a critical global concern. The European Union (EU), widely regarded as a leader in digital regulation, has developed an AI governance architecture that includes the AI Act, Ethics Guidelines for Trustworthy AI, and multiple stakeholder platforms.<br /><strong>Aims: </strong>This article examines how expert networks and national governance models from France, Germany, and South Africa contribute to the European Union’s artificial intelligence (AI) governance through conceptual, institutional, and procedural spillover.<br /><strong>Methodology: </strong>The research employs comparative case study analysis, drawing on policy documents, ethical guidelines, expert reports, and process tracing to track how national frameworks migrate into EU deliberations. The theoretical framework integrates spillover theory with multi‑level governance, norm diffusion, and epistemic community perspectives.<br /><strong>Discussion: </strong>The central question is how transnational actors influence EU regulation via mechanisms such as normative transfer, expert mobility, and platform convergence. The study hypothesizes that EU AI governance is increasingly co‑constructed through multidirectional spillover, in which norms and ethical frameworks from both the Global North and Global South are adapted and embedded into supranational regulation.<br /><strong>Findings:</strong> South Africa’s “Fair AI” framework, France’s participatory ethics inquiry, and Germany’s strategic critique each shape EU debates on trustworthy AI, accountability, and regulatory experimentation. Together, these cases indicate that EU governance functions not as a closed, top‑down system but as a porous and adaptive architecture responsive to external influence.<br /><strong>Conclusion:</strong> The expert communities across regions contribute to the co‑production of global AI standards, and that ethical pluralism is emerging as a significant feature of supranational regulation. https://jcss.ut.ac.ir/article_105007_ccfa3ef6f43ce023b23eb3e7c0d8f5d9.pdf