Is AI (Failure) Here to Stay? An Examination of Stability Attributions, Customer Responses, and Repurchase Intentions
Artificial intelligence (AI) technologies are increasingly integrated into customer-facing financial services, particularly within the wealth management industry where firms rely on AI-driven systems for portfolio recommendations, fraud detection, customer communication, and financial advising support.While these technologies offer efficiency, they also introduce new forms of service failure that may alter how customers evaluate firms regarding trust, satisfaction, and frustration.Existing service failure and recovery literature has extensively examined customer responses to human-caused failures; however, comparatively less research has explored how consumers cognitively and emotionally process service failures caused by AI systems.More specifically, little attention has been given to whether customers perceive AI failures as temporary incidents or as indicators of broader systemic instability within the firm.This study seeks to examine customer responses to human-and AI-caused service failures in the wealth management industry through the lens of Cognitive Appraisal Theory (CAT), Construal Level Theory (CLT), and Attribution Theory.Specifically, a research model is developed to investigate how perceptions of service failure severity influence attributional stability, defined as the extent to which consumers perceive a failure as ongoing, permanent, and likely to recur over time.The model further examines how stability attributions affect trust, firm satisfaction, frustration, and repurchase intentions following service failures involving either human or AI error.
Authors
- En Mao (ORCID: https://orcid.org/0000-0003-4390-4002)
- Kaitlin Gravois
- Adrien Maught
Institutions
- Louisiana State University (US)
- Nicholls State University (US)
Publication Details
- Journal
- DigitalCommons - Kennesaw State University (Kennesaw State University)
- Published
- 2026-11-20
- Primary Topic
- Customer Service Quality and Loyalty
- Type
- article
- Field-Weighted Citation Impact
- 0.00