It is suggested that technical version succession does not necessarily amount to effective replacement on the user side, and user experience can provide important information for identifying post-deployment impacts and should be incorporated into lifecycle evaluation and decision-making.
Abstract
AI model lifecycles are commonly understood as a series of technical and organizational processes. Yet once a model enters sustained use, subsequent changes can also affect established user practices and user value. Using the Keep4o movement around GPT-4o as a case, this study examines post-deployment AI model lifecycle issues from the user side. We collected 61,846 public original posts on X from August 2025 to March 2026 and, using a systematically developed coding framework and LLM-assisted content analysis, analyzed discussion themes, users'reasons for wanting to keep GPT-4o, and the specific claims they made. Findings show that the Keep4o discussion extended well beyond continued access to the model itself. It covered concrete experiences of use, model behavioral characteristics and how they changed, and management issues across different stages of the model lifecycle. Reasons for keeping GPT-4o reflected interactional and relational value formed through long-term use, as well as judgments about the adequacy of replacement and the reasonableness of related decisions. The corresponding claims further reflected users'specific expectations for model lifecycle arrangements and governance. Overall, the call to"keep GPT-4o"brought together different judgments about user value and governance concerns. These findings suggest that technical version succession does not necessarily amount to effective replacement on the user side. Post-deployment AI model lifecycle management therefore needs to consider whether established user value can be carried forward and how model changes affect actual use. This study thus provides user-side empirical evidence for AI model lifecycle management. It further shows that user experience can provide important information for identifying post-deployment impacts and should be incorporated into lifecycle evaluation and decision-making.
Interviews with sixteen early-adopter software professionals who integrated LLM-based tools into their day-to-day work in early to mid-2023 offer actionable implications for developers, organizations, educators, and tool designers seeking to integrate LLMs responsibly into professional software practice.
Benyamin T. Tabarsi, Heidi Reichert, Sam Gilson et al.· Empirical Software Engineeri...· 21 citations· ⚡1
This study examines how organizations integrate generative artificial intelligence (GenAI) into everyday business processes and workflows, and how governance practices are emerging alongside such integration. While prior research has focused on task‐level performance effects or adoption intentions, less evidence examines how GenAI is embedded in organizational workflows and governed in practice. The study adopts an exploratory, descriptive research design based on survey data collected from 120 practitioners across organizations of varying sizes and sectors. The survey captures reported patterns of GenAI use, workflow integration, governance practices, safeguards, and outcomes, and the analysis relies on descriptive statistics and exploratory comparisons across organizational contexts. The findings indicate that GenAI use is widespread in the sample, but integration into business processes is incremental and localized rather than deeply embedded. Many organizations rely on third‐party tools and user‐driven experimentation, while deeper enterprise‐level integration and formal governance arrangements are less frequently reported. The results show uneven development between workflow‐level use, technical embedding, operational transformation, and governance formalization. Operational safeguards appear more common than enterprise‐wide governance structures. Respondents report improvements in efficiency, productivity, and decision‐making alongside concerns about privacy, security, bias, and explainability, indicating experimentation. This study provides cross‐organizational, practitioner‐reported descriptive evidence on how organizations are integrating GenAI into business processes and how governance practices are evolving alongside its use. The findings extend current empirical understanding of how organizations are reporting early GenAI use and governance in practice by moving beyond isolated tasks and adoption intentions to examine workflow‐level integration across organizations.
Shirin Hasavari, J. Zaveri· Knowledge and Process Manage...· 0 citations
This paper presents a transformation model designed to assist organizations in initiating their journey into Metaverse Use Cases. In order to have a human centric approach, the model articulates a structured path along a multitude of questions which shall guide organizations to start a successful Use Case in the Metaverse. The questions are categorized in five different areas. Organization goals, organization prerequisites, requirements to the Metaverse and the technology around it, implementation of the Use Case itself and a learning phase. It encourages the organisations to emphasize on getting started right away and integrate feedback loops for learning from the current development at every step of the journey. To help organizations to find those collaboration opportunities, a survey was derived based on the transformation model. It aims to validate the structured approach of the model and guide organizations through four of the five different transformation areas. At the end organizations receive recommendations for support for their Use Case through companies, academia, public of non-profit projects and other means. Building on these insights, the transformation model helps to democratize access to Metaverse initiatives by providing a practical roadmap that can be adopted across diverse organizations and industry sectors. By foregrounding cross-sector partnerships, the approach reduces barriers to entry and accelerates capability development. The accompanying survey further strengthens this framework by translating qualitative guidance into actionable recommendations for industry, academia, and public or nonprofit partners.
Franz Falkenau, Peter Schrader, Benjamin Wingert et al.· AHFE International· 0 citations
This study investigates the adoption and utilisation of AI Builder in the Microsoft Power Platform; the focus will be on the socio-technical barriers affecting the successful uptake within organisations. A qualitative approach utilizing a single-case study of documented AI Builder implementation at the Agriculture and Horticulture Development Board (AHDB) is used along with existing literature. A thematic analysis was performed to analyse the implementation process and the critical organizational and technical aspects of AI Builder adoption. As shown, governance, citizen development, and organizational preparedness were found to be major themes for effective implementation, and good data governance, collaborative efforts across teams, change management, and user trust are enablers for successful, sustained adoption. The AHDB case shows how AI Builder, when used with Power Apps, Power Automate, Power BI and Dataverse, can lead to better coordinated workflows, single data points and a more efficient organisation. The study demonstrates that the successful deployment of AI Builder is contingent more on organizational readiness and accountability than technological maturity. These findings offer an informed blueprint to organizations adopting low-code AI platforms and contribute to the ongoing body of literature in AI democratization and socio-technical implementations.
P. Vutla, Triveni Yenugu· Journal of Information Techn...· 0 citations
Despite high expectations about the benefits of data analytics (DA), our understanding of the mechanisms that drive value creation from the implementation of DA in firms remains incomplete. While some studies have suggested that innovation prowess, human resources and operational capabilities are pieces of the puzzle of DA effectiveness, how these pieces fit together remains an open question. In this research, we study the mechanisms that explain value creation through DA implementation in firms, focusing on the use of DA in operations management.
We empirically address this question by analyzing firm-level data from the European Company Survey (ECS) 2013 (19,470 managers) and 2019 (18,616 managers and 1,848 employees), covering 28 European countries. We conduct four sets of robustness tests: (1) comparisons of results among the three subsamples (2013-manager, 2019-employee and 2019-manager); (2) multiple endogeneity checks; (3) alternative model specifications and (4) alternative model assessments.
We present evidence showing that DA impacts firm performance primarily through the enhancement of capabilities for process and product innovation. Further, in line with sociotechnical systems (STS) theory, our moderated mediation analyses show that the value created from DA is contingent upon the presence of employee involvement practices including empowerment, development and team orientation.
This research integrates DA, innovation capabilities and employee involvement within an STS framework to explain firm performance, using three large-scale, cross-sectional European datasets, going beyond the conceptual and empirical examination of DA in prior studies.
B. Lameijer, J. De Mast, Leopoldo Gutierrez et al.· International Journal of Ope...· 0 citations
The findings show that AI-CBM transitions unfold through recursive cycles of experimentation, validation, and recalibration, as developments in one dimension expose misalignments in data maturity, governance arrangements, and ecosystem coordination.
M. Kowalski· Journal of Environmental Man...· 0 citations