Understanding AI service failures: insights from attribution theory
Wenqi Zhang et al.
Abstract
Purpose Despite the rapid proliferation of AI services, scholarly examinations on the AI service failures remain sparse. This research aims to investigate AI service failures via the lens of the attribution theory. Design/methodology/approach Drawing on the attribution theory, this research proposes a framework on AI service failures. Leveraging a large scale of negative online reviews of AI services and an annotation survey, our research examined consumers' attributions of AI service failures by locus of causality, controllability and intentionality. Findings AI service failures can be classified by locus of causality into AI Algorithm Failures, Delivery Medium Failures and Commercial System Failures. Failures with distinct loci of causality can be associated with different levels of perceived controllability and intentionality; these perceptions are further associated with service evaluations. Practical implications The findings offer important implications on the management of AI services and relevant strategies to effectively mitigate and handle AI service failures. Originality/value Our research proposes a novel theoretical framework on AI service failures. Dovetailing attribution theory and service system thinking, our research reveals the attribution process of AI service failures toward diverse entities within the complex AI service system.
Evidence weight
Balanced mode · F 0.40 / M 0.15 / V 0.05 / R 0.40
| F · citation impact | 0.50 × 0.4 = 0.20 |
| M · momentum | 0.50 × 0.15 = 0.07 |
| V · venue signal | 0.50 × 0.05 = 0.03 |
| R · text relevance † | 0.50 × 0.4 = 0.20 |
† Text relevance is estimated at 0.50 on the detail page — for your query’s actual relevance score, open this paper from a search result.