Editorial: Service innovation in the AI era

The service sector dominates the economy, with services – encompassing everything from healthcare and education to finance, entertainment, hospitality, transportation, government, information technology and much more – now making up most GDP and employment globally (OECD, 2017; World Bank, 2023). In this context, our understanding of economic exchange has shifted from focusing on the production of tangible output to focusing on value co-creation within business and social systems. On this view, value is not something embedded in a product at a factory; rather, value is co-created when skills and knowledge are applied for mutual benefit at the time of and in the context of use (Vargo & Lusch, 2004). Whereas historically, models of service innovation treated services as intangible products (Gallouj & Weinstein, 1997; Miles, 2010; see also Barrett, Davidson, Prabhu, & Vargo, 2015), service innovation is now best viewed as institutionalization of new practices that change how value is created (Vargo, Wieland, & Akaka, 2015) by reconfiguring roles and relationships among economic actors (Normann & Ramirez, 1993) and effectively resetting institutional rules that govern business and social interactions among them (Koskela-Huotari, Edvardsson, Jonas, Sörhammar, & Witell, 2016). This essay explores how service innovation operates as a kind of institutional work and how digital technologies, particularly artificial intelligence (AI), accelerate innovation by enabling novel configurations of actors and resources in service systems (Lusch & Nambisan, 2015).Reconfiguring the roles of actors in a service system changes how value is created in that system, and if orchestrated correctly, can increase the value created overall (Normann & Ramirez, 1993). Service innovations often succeed by shifting the boundary of actions and responsibilities among actor roles. For instance, IKEA innovated with customer-assembled, flat-packed furniture, radically reconfiguring its service system by shifting transportation and assembly from company to customer. Similarly, McDonald's revolutionized food service not just through internal optimization and technology but also by training customers how to order and clear their own tables, among other changes. By shifting who does what, these companies unbundled traditional service models and created more overall value in the process by reducing costs while offering customers lower prices and immediate availability (see also Campbell, Maglio, & Davis, 2011).Institutions establish the regulations, normative expectations and cultural dimensions of social life (Scott, 2014). For a service innovation to take hold, it must do more than offer a new technical capability; it must alter the institutional arrangements that determine how value is assessed and exchanged. Thus, innovation requires institutional work in which actors break existing rules, make new rules and maintain these new structures until they become the accepted standard (Koskela-Huotari et al., 2016). Central to this institutional view is the reconfiguration of roles and relationships within a service ecosystem. For example, in the healthcare service system, the shift from provider-centered to patient-centered care represents a service innovation that was not primarily a technological change but a relational one: the patient's role shifts from passive recipient of treatment to active partner in wellness and the physician's role shifts from authoritative prescriber to collaborative partner, changing the institutional norms and establishing new practices for information sharing (Patrício et al., 2020). Similarly, the sharing economy platform Airbnb did not create new housing stock; it reconfigured the role of the homeowner from private resident to service provider and established new institutional rules (trust mechanisms) to govern relationships between strangers (Lusch & Nambisan, 2015).Digital technology facilitates institutional change and role reconfiguration through resource liquefaction, decoupling information from its physical container, allowing it to be dematerialized, manipulated and easily shared across networks (Lusch & Nambisan, 2015; Normann, 2001; Vargo, Fehrer, Wieland, & Nariswari, 2024). For example, physical lodging resources are static, but when liquefied into information on a digital service platform (such as Airbnb), they become accessible to a global market (Lusch & Nambisan, 2015). Such platforms make resources accessible and allow for rapid matching of needs and capabilities (Normann, 2001). And by liquefying physical assets, platforms can also allow firms to operate without operations, coordinating value creation without owning any underlying physical assets (Breidbach et al., 2018, 2022).As service systems become increasingly infused with digital technology, they may evolve into smart service systems capable of sensing and responding to their environment by connecting people, devices and infrastructure (such as sensors in a smart city or mobile phones in a transit network); collecting data generated by these connected sensors; computing and processing data, identifying patterns and generating predictions (such as forecasting traffic congestion); communicating insights back to system participants, informing them of optimal choices; co-creating value by using this information to help integrate resources and controlling the system to take autonomous action based on the data, optimizing operations in real-time (Lim & Maglio, 2018). For instance, smart traffic systems can analyze real-time data to manage signals, and healthcare platforms can use data from fitness trackers to provide personalized coaching (Maglio & Lim, 2016). In each case, data analytics liquefies a physical resource – traffic flow or health status – enabling new forms of value creation among system actors.The integration of AI and robotics represents the frontier of role reconfiguration in service systems, presenting a design challenge (Wirtz et al., 2018). AI provides powerful resources that can learn, reason and act, with increasing capabilities and autonomy (Huang & Rust, 2018). Effective working relationships between humans and AI systems make effective tradeoffs between automation and augmentation (Yee et al., 2025). Automation relies on technology to replace human labor, typically for transactional tasks that are routine, repetitive and easily codified, standardizing processes and reducing costs (Rouse & Spohrer, 2018). But automation can fail in unanticipated situations that require flexibility and judgment (cf. Kandogan, Maglio, Haber, & Bailey, 2011). By contrast, augmentation uses technology to support and amplify human capabilities, particularly in tacit roles that require complex thought, judgment, empathy and communication. For example, AI may automate the data-intensive tasks of analyzing market trends or monitoring system performance, while human workers focus on evaluating complex products or crafting detailed strategies (World Economic Forum, 2023; Yee et al., 2025). This suggests that sustained competitive advantage lies in mastering complexity through augmenting human capabilities rather than optimizing routine transactions (see also Manyika, Johnson, & Yee, 2005). For example, digital twins – AI-based representations of individuals that can act on their behalf – might handle routine interactions (such as scheduling, basic correspondence or information retrieval; see Spohrer, Maglio, Vargo, & Warg, 2022; Spohrer, 2024), liquefying the individual's presence and expertise. The effectiveness of an augmented workforce depends on establishing robust frameworks for human-machine teamwork (Kunz, Sajtos, & Flavián, 2025; Mortati & Freitas, 2025).Reconfiguration of service systems through AI is not without its perils. The data-driven core of AI-enabled service systems raises issues regarding accountability and fairness (Breidbach & Maglio, 2020). And as AI advances quickly, it seems likely that governance frameworks will fail to keep pace; for instance, the potential of conscious AI or highly autonomous service robots may require new institutional structures to manage the rights of digital workers (Breidbach et al., 2025).