Cognitive-Technology Implementations’ Transformative Dynamics: Socio-Technical Implications for Knowledge-Intensive Work in the Financial-Accounting Domain

The rise of cognitive technologies such as artificial intelligence (AI) with learning capacity and of advanced rule-based automation systems have in recent years called humans' dominance in the domain of cognitively oriented knowledge-intensive work into question. Proliferation of cognitive technologies, propelled by rapidly expanding interest in applying AI to such knowledge work, has opened a wide spectrum of AI development and implementation projects, with the variety and quantity of cognitive-technology initiatives ballooning accordingly. Therefore, AI's presence in knowledge-work organizations is growing. As it comes to rival rule-based automation systems (which follow a set of premeditated steps), the application of AI is enabling both automation and augmentation of a broader range of work tasks, thanks to capabilities of learning. Novel cognitive technologies facilitate offloading of repetitive tasks, even highly complex ones, via automation that presents possible solutions to the particular task at hand; furthermore, they can stimulate insight into the work through the information with which they augment humans' decision-making. However, organizations need to tread carefully when deploying these technologies: alongside beneficial humanistic and instrumental outcomes, such technologies can lead to disruptive dynamics that may produce detrimental outcomes for individuals and organizations alike, such as deskilling of the overall socio-technical system. Responding to these concerns, three qualitative studies were conducted to enrich socio-technical research related to cognitive-technology implementations' implications for knowledge intensive work. Essay 1 examines the transition toward human–AI work and the subsequent transformation of practices through shifts in routines' dynamics and changes in work roles. Essay 2, in turn, examines negative elements: how exploiting cognitive automation encourages erosion of knowledge workers' skills and the effects of this skill erosion at the level of organizations, and the third essay explores why AI system development and deployment projects face delays and complications. The research deepens understanding of the evolving dynamics among the practices employed in these systems' development and implementation. By addressing key phenomena linked to these phenomena, the dissertation contributes to both theory- and practice-oriented work on topical issues of developing and implementing cognitive technologies for utilization in knowledge-intensive work processes. Offering insight as to the patterns that can lead to beneficial/detrimental outcomes, the research serves the aim of expanding understanding of cognitive technology's socio-technical implications.