Betrayed by AI: How perceived betrayal by a virtual assistant affects consumers’ purchase intentions for recommended products
[Display omitted] •Consumers can experience betrayal in user relationships with AI-based VAs.•An act of betrayal by a VA can lead to psychological discomfort for consumers.•Feeling betrayed by a VA can diminish consumers’ relational closeness with a VA.•Betrayal can make consumers less likely to fol...
Saved in:
Published in | Journal of business research Vol. 185; p. 114940 |
---|---|
Main Authors | , , |
Format | Journal Article |
Language | English |
Published |
Elsevier Inc
01.12.2024
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | [Display omitted]
•Consumers can experience betrayal in user relationships with AI-based VAs.•An act of betrayal by a VA can lead to psychological discomfort for consumers.•Feeling betrayed by a VA can diminish consumers’ relational closeness with a VA.•Betrayal can make consumers less likely to follow recommendations made by the VA.
AI-powered virtual assistants (VAs), such as Amazon’s Alexa, have transformed consumers’ interactions with technology. Consumers develop relationships with VAs, a phenomenon that has proven beneficial to firms. By employing the “computers are social actors” (CASA) paradigm, we examine a potential vulnerability in the consumer–VA relationship. Users’ relationships with VAs may expose firms to negative consequences when consumers perceive that their VA has betrayed them. Across three studies, we demonstrate that VA betrayal reduces consumers’ purchase intentions for products later recommended by their VA. The VA’s betrayal generates psychological discomfort for the user, which amplifies the user’s perceptions of betrayal and reduces their feelings of closeness to their VA. This process reduces users’ purchase intentions toward products recommended by their VA. VA developers should consider design features that recognize and repair the user–VA relationship after a perceived betrayal. |
---|---|
ISSN: | 0148-2963 |
DOI: | 10.1016/j.jbusres.2024.114940 |