Governing AI Innovation Ecosystems: A Triple Helix Network Analysis of Global AI Research Hubs

• In Event: Governance Ecosystems: Innovation, Institutional Design, and Policy Outcomes

Abstract

As artificial intelligence (AI) emerges as a general-purpose technology (OECD, 2025), governments are investing in dedicated AI research hubs, yet similar investments have produced markedly different ecosystems. Drawing on Etzkowitz’s Triple Helix framework, this study asks how the institutional configuration of a hub shapes the university-industry-government relations it organizes. It compares three hubs: the German Research Center for Artificial Intelligence (DFKI), a non-profit company (GmbH) of universities, firms, and government; the Alan Turing Institute (ATI), a university-led national institute; and the RIKEN Center for Advanced Intelligence Project (RIKEN AIP), a government-led research center. Combining three layers of relations - formal partnerships, talent circulation, and co-publication networks (2017-2026) - with collaborations among each hub’s leading partners outside the hub and a supplementary analysis of patent-to-paper citations, the study finds that each configuration leaves a distinct network signature on a common university base. ATI connects the most diverse set of sectors; DFKI draws the most industry partners and shows the strongest technological uptake; and RIKEN AIP is the most centralized and domestically anchored, yet bridges partners that do not otherwise collaborate. Across all three, declared partnerships mirror the hub’s configuration, whereas talent circulation and co-publication converge on academic and international partners. The study contributes to policy scholarship on innovation systems by reframing AI innovation as a function of governance structures and network coordination. It also offers practical insights for policymakers designing AI hubs.

Author

• Saebom Jin, Korea Advanced Institute of Science and Technology
  Presenting Author