Recently, Machine Learning (ML) has attracted attention as a technology for organizations to automate processes and innovate new services. Data-driven technologies and analytics are also receiving increased interest among IS scholars. However, whereas much research has been concerned with the potential for ML technologies, less attention has been paid to empirical accounts of how ML is applied in organizational contexts. Also, ML tends to be portrayed as a stand-alone technology and not as part of a larger arrangement of data sources, applications and systems. Against this backdrop, this paper investigates the challenges of introducing ML in a governmental organization. In so doing, we report from an ongoing case study of a large governmental organization in Norway attempting to use ML as a means for new service innovation and automation of case handling. We find that ML have numerous challenges related to utilizing, evaluating and integrating various data sources, need of adopting to a ML-friendly infrastructure, as well as legal challenges related to GDPR and sensitive data in general. We contribute by illustrating how an information infrastructure perspective is helpful for analyzing and theorizing ML in organizational contexts. We also identify further research questions for the latter part of the study.



When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.