Anna Essén

The proliferation of generative artificial intelligence (GenAI) in the workplace raises critical questions on how to regulate these technologies within organizations. GenAI technologies generate excitement but also present significant ethical and legal issues, compelling organizations to self-regulate in the absence of comprehensive governmental regulations. Although the emerging field of regulation and information technology (IT) provides a foundation for exploring rulemaking in relation to new technologies, the unique trajectories of GenAI remain underexamined. Through a field study of a leading European business school, we trace how an organizational group without technological expertise assumed responsibility in this context. Our preliminary findings reveal how rulemaking around GenAI evolved through the contingent interactions between practices, rules, and technology, where rule-makers transitioned from preventing unwanted uses by “copying” existing regulations to crafting distinct, tailored rules and finally facilitating learning to promote new, wanted uses. Throughout this journey, the rule-makers grappled with the challenge of regulating an emergent technology for which they lacked initial expertise and the envisioned usages and risks were constantly shifting.

Incumbent firms typically face significant risk of losing the relevance of their physical core when facing industry disruption driven by digital technologies. Existing literature emphasizes a digital first approach, whereby firm offerings are fundamentally redeveloped from a digital point of view, from the point of conception. While this prescription can help accelerate innovation, it does not tell us how incumbents might safeguard the relevance of their traditional physical core resources when going digital first. This is important, since major discontinuities in strategic repositioning, while often celebrated in digital innovation and transformation literature, create significant risks to firm survival. To this end, we conduct a grounded analysis of a European automotive firm’s innovation journey over an eight-year period. We contribute to the digital innovation and transformation literature by developing a process model explaining how a digital first approach can be employed in a way that also safeguards the physical core.

Research has explored how algorithmic processes reshape individual tasks in organizations. But less work has explored their implications on the wider service ecosystem. Using data about the “DAS” algorithms in the Swedish rheumatology setting between 1995 and 2019, we examine how the use of algorithms can reconfigure service delivery across the service ecosystem. We identify two algorithmic translation processes that enabled the incorporation of consumer self-evaluations in the multiple service nodes of the studied ecosystem. These processes implied new forms of multilateral surveillance, which had implications on the service delivery and the values that were generated by the ecosystem. By showing how concealed algorithms can make consumers’ voice watch-able and action-able across multiple nodes in the ecosystem, our work challenges traditional conceptualizations of algorithmic black-boxing as reducing human agency. We further theorize algorithmically mediated multi-actor surveillance, in which service nodes are both being watched and watching others, thus also challenging the view of surveillance as involving one “invisible” controlling many visible actors. These results open up new avenues of inquiry about the role of algorithmic processes in enabling new forms of multilateral surveillance that in turn condition value creation in service ecosystems.