Institutions and society face profound challenges in navigating volatile environments and complex, ill-defined problems. Diverse teams of experts are often assembled to address these grand challenges. Collaboration in knowledge-diverse teams is indispensable to tackling these challenges, but integration of knowledge across specialties is difficult. Designing interventions to improve team performance is complex because of the specific nature of each collaboration. We developed a theoretically driven scaffolding intervention for early-stage biomedical collaborations and tested our conceptual model in a field-based quasi-experimental design. Our sample consisted of 26 newly formed biomedical research collaborations involving 127 scientists. Teams that received the scaffolding intervention showed greater knowledge integration than those in the comparison group, and this increase in integration was associated with stronger downstream performance, including more grants secured, more publications produced, and higher journal impact. We also find that the effectiveness of the scaffolding intervention was mediated by the extent to which collaborators could coordinate with one another to integrate their diverse expertise.
Organization development (OD) seeks sustainable change, yet its theoretical toolkit remains underspecified regarding the normative mechanisms through which organizational context shapes change-supportive behavior. This article introduces Publicness-Activated Change Theory (PACT), positioning perceived organizational publicness—the extent to which employees experience their organization as attentive to the public interest—as a potential cross-sector institutional change mechanism. Drawing on institutional theory, the logic of appropriateness, and internalization theory, PACT conceptualizes perceived publicness as normative change infrastructure through which public values may shape judgments about appropriate behavior. The framework makes three contributions: it theorizes perceived publicness as an informal institutional context for change; introduces publicness-values misfit as a diagnostic explanation for resistance; and presents a three-stage Diagnose–Align–Leverage model grounded in the Publicness Perceptions Scale. Together, these contributions provide a testable framework for examining when publicness can be leveraged, when alignment is required, and how its effects compare with established OD mechanisms.
Facing pressure to change, organizations are increasingly adopting data-driven change management technologies (DDCMTs), digital systems that collect and analyze change-relevant data to inform and support change management. Due to their novelty and equivocality, users’ understanding and use of DDCMTs hinge on how they make sense of these technologies, a process that has received little empirical attention. Therefore, we present an interdisciplinary multiple-case study of three DDCMTs. Our analysis shows that users’ understandings of DDCMTs center on transparency and control but vary considerably. Further, we illustrate how, beyond their potential benefits, the digital characteristics of DDCMTs—editability, autonomy, and interactivity—generate equivocality that allows for multiple and divergent understandings. Thus, we advance research on digitally supported change management by explaining how users develop multifaceted and ambivalent understandings of DDCMTs, providing an important step toward understanding their use. We also offer practical guidance for the development, implementation, and effective use of DDCMTs.
Agentic artificial intelligence (AI) systems are entering the workplace at an unprecedented pace, making the associated organizational change a unique challenge for senior leaders. Traditional change models assume controllability and sequential implementation. By contrast, agentic AI operates autonomously, and adoption often emerges through employee exploration, rendering established blueprints a limited guide. Conceptualizing top-down steering and bottom-up enactment as independent dimensions rather than opposing choices, we outline four prototypical approaches to agentic AI change: (1) light-touch drift, (2) grassroots experimentation, (3) command-and-scale, and (4) integrative change. Rather than pinpointing a superior strategy, we argue that senior leaders must continuously diagnose and rebalance their approach to fit their context. We call for treating agentic AI change not as a one-time implementation challenge but as the ongoing orchestration of an evolving portfolio of top-down and bottom-up processes. We close with directions for future research on the enactment, temporal patterns, and governance of agentic AI change.
The electrical grid is a prime example of a vast system of interconnected entities that are both empowered and challenged by artificial intelligence (AI). AI is entering this sector in two ways: as a deliberate instrument of AI-led change, enabling better forecasting, grid optimization and increasingly autonomous control approaches; and as an uninvited force of AI-generated change, driven by the enormous, spatially concentrated and rapidly evolving electricity demand of AI data centers. This demand does not arrive as an initiative or technology to be adopted, but rather as a massive load interconnection request. AI-generated change challenges some of the fundamental principles of existing literature on organizational change, such as the authorship of change, the composition of the changing unit, and the ability of the involved parties to share a joint agenda. We use the interplay between AI and the electricity sector to illustrate how authority over change is shifting toward those controlling allocation, interfaces, and architecture. The paper proposes avenues for further research into organizational change in the AI era.
