Co-creating AI experimentation ecosystems: Designing business and governance models for Smart Cities

As the European Union advances its AI regulatory agenda, cities are becoming critical arenas for testing and validating emerging technologies under real-life conditions. This paper examines how AI Testing and Experimentation Facilities (TEFs), a hybrid form between testbeds, sandboxes and Living Labs, can be organized as distributed, node-based ecosystems to support responsible innovation in urban environments. Drawing on insights from the Digital Europe project CitCom.AI, spanning over a dozen TEF nodes across Europe, we explore how cities and regional actors can implement shared experimentation infrastructures while retaining local autonomy and regulatory alignment. The paper builds on a dual theoretical framework combining Multi-Level Governance and Business Ecosystem Theory. Through mixed methods, we investigate how governance structures, business models, and service specializations are evolving across the network. Preliminary findings suggest a set of core design principles for TEFs; role complementarity over service replication, distributed governance with aligned ethical standards, service interoperability across nodes, stakeholder-anchored validation, and regulatory readiness as a built-in service. These principles offer a blueprint for structuring TEFs that are adaptable to different urban contexts yet coordinated through a shared European vision. Our contribution supports the development of scalable, trusted, and citizen-centred AI innovation frameworks across smart cities in Europe. We also contribute to the literature on sustainable business models for Living Labs and to the literature on Living Labs for emerging technologies.