Abstract
As Artificial Intelligence (AI) is becoming more prevalent in everyday work settings, the Information Systems (IS) discipline is perfectly poised to study the sociotechnical repercussions of algorithmic decision-making part of expert’s knowledge work. As expert know-how is tacit and socially situated, there are difficulties in capturing the nuances of the emerging human-machine collaborations. In this article, we review how epistemic uncertainty is evident in experts’ decision-making practices when using AI tools. Building on rich primary studies, a qualitative evidence synthesis (QES) approach was used to bring together and analyze relevant literature on this topic. Our findings unravel sources of expert uncertainty, point to strategies experts use to cope with uncertainty, and what attitudes experts have towards uncertainty when making complex decisions with AI.
