Insights from People's Experiences with AI: Privacy Management Processes.

Given the lack of interview-based research in the Artificial Intelligence (AI) literature, we conducted 15 semi-structured interviews about people’s experiences with AI. From those interviews, many important themes emerged and we focused this paper on people’s experiences with privacy violations. We used the Communication Privacy Management Theory (CPM) to analyze distancing behaviors (avoidance or withdrawal) resulting from Boundary Turbulence (privacy violations) with people’s voice-activated phones and Google Search. Through this analysis, we found evidence that distancing behaviors may 1.) be able to transfer from one device to another; and 2.) not depend on the trustworthiness of a particular device or application company. This paper concludes with recommendations for applying the CPM framework more intentionally and rigorously to people’s experiences with voice-activated systems.