Governing the Data Backbone: Designing and Deploying Multi-Actor Nodes for AI-Driven Anti-Corruption Policy

Data holds significant potential for anti-corruption efforts, yet it remains dispersed across multiple public institutions and private actors. The key public management challenge lies in developing and administering dedicated data nodes — institutional hubs through which such data can flow, be analysed, and acted upon, ideally via AI systems. These nodes may be organised in a range of models, from simple coordination mechanisms to advanced, real-time networks, to networks that enable predictive analytics and targeted preventive actions in projected corruption hotspots. This paper pursues two aims. First, it outlines a conceptual framework for the design and evolution of data node infrastructures to support AI deployment in anti-corruption policy. The framework envisions a continuous progression from simple coordination between a small number of data-holding institutions to more sophisticated, integrated systems enabling not only detection of past corruption patterns but real-time interventions to stall ongoing corruption attempts as well as predictive interventions aimed at reducing corruption in the future. Second, the paper empirically maps how five to ten countries, primarily in Europe, have implemented such models in practice, assessing both their current stage of development and future potential. The core contribution lies in shifting the focus from the software capability of AI in corruption detection to the infrastructural and governance dimension — specifically, how public management can organise, sustain, and coordinate multi-actor data ecosystems to enable effective AI use in the fight against corruption.