Work

Substitutive and Augmentative Delegation in GenAI ‐Assisted Shopping: Perceived Benefits, Constraints, and Consumer Tensions

Year: 2026

Type: article

Abstract: ABSTRACT Generative artificial intelligence (GenAI) is reshaping consumer shopping by introducing new forms of decision support while also creating new tensions. Using the technology paradox perspective as a sensitizing theoretical lens alongside delegation theory, this study examines how perceived GenAI benefits (process efficiency, personalization, creative discovery, and immersive experience) and constraints (hallucination potential, privacy intrusiveness, algorithmic opacity, and information biases) are associated with performance ambiguity, emotional ambivalence, and autonomy threat through substitutive and augmentative delegation‐related use patterns. Using survey data from 438 Chinese consumers, partial least squares structural equation modeling (PLS‐SEM) showed that process efficiency was negatively associated with substitutive delegation and positively associated with augmentative delegation. Perceived hallucination potential was also negatively associated with substitutive delegation, whereas perceived information biases were positively associated with it. Substitutive delegation was positively associated with all three consumer tensions, while augmentative delegation was negatively associated with emotional ambivalence and autonomy threat but was not significantly associated with performance ambiguity. Furthermore, a separate stage‐based comparison analysis with 400 additional respondents revealed differences among four purchase‐stage samples. Emotional ambivalence was higher in the discovery and comparison samples than in the purchase and post‐purchase samples, whereas performance ambiguity and autonomy threat differed only in selected stage comparisons. The findings extend understanding of consumer responses to GenAI‐assisted shopping and offer implications for designing more transparent, controllable, and consumer‐centered GenAI shopping tools.

Source: Journal of Consumer Behaviour

DOI: 10.1002/cb.70238

Authors: Kexin Meng, Mousumi Singha Mahapatra, Jing Jian Xiao

Institutions: Xiamen University of Technology, Institute of Public Enterprise, University of Rhode Island

Language: English

Topic: AI in Service Interactions

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: bronze

Funders: Xiamen University of Technology, Xiamen University

OpenAlex ID: W7204233963

FWCI:

Cites: 120

Cited by:

Related to: