Artificial intelligence (AI), like many technological innovations before it, promises to revolutionize organizations. However, implementing AI in organizations is not as simple as it may appear. This exploratory research aims to unearth barriers to the implementation of AI in organizations. The methodology is based on a ranked-order Delphi study with 18 AI experts. By comparing our results with previous research on barriers to implementation of other information systems and to conceptual and practitioner literature on AI implementation, our findings underscore specific AI implementation challenges for organizations. Barriers to AI implementation fall under three main categories: (1) a lack of organizational capabilities related to data; (2) a lack of individual competencies related specifically to AI; and (3) generic implementation barriers previously observed in implementation research that persist with this innovation.

While the Agile-Scrum (scrum) framework has specific guidelines, these guidelines are often adapted by practitioners. This research aims to understand how scrum changes in practice and how these changes impact various aspects of project success. Through interviews with representatives from 11 organizations who use scrum for software development, we found variability in the application of the guidelines, namely, that only a small number of guidelines are systematically followed, and that some guidelines are rarely followed consistently. Examining these method deviations and mapping them to specific dimensions of project success, four patterns emerged. Further, we uncovered practices that are often followed but were not part of the original Scrum guidelines, including how organizations scale scrum projects. These insights into how scrum is used in practice can help industry professionals determine how to best adapt scrum. They also serve as a promising agenda for research on the application of the scrum framework in industry.