Abstract: Empowered by rapid developments in artificial intelligence (AI), customers frequently encounter AI-based services across multiple domains. For instance, an AI assistant can offer investment advice and then extend its services to provide legal consultation.
Eleven studies (and two supplemental studies) document that AI (vs. human) service providers better retain customers’ trust when they extend their service offerings from an established domain to a novel domain, a pattern that holds across various populations, domains, as well as measures of trust and behavioral outcomes. This AI advantage emerges because customers attribute the success of AI (vs. human) service providers in the established domain to data (vs. experience), which is construed as more generalizable across domains. Consequently, customers grant AI a generalizability premium. We document that (i) describing human service providers as relying on generalizable data or (ii) AI providers as relying on domain-specific data attenuates the effect. While the generalizability premium of AI is robust across various service extensions, it is smaller for highly similar or dissimilar domain extensions. These findings advance our understanding of how customers construe the success of service providers and offer practically relevant insights on how companies should manage and communicate service extensions to retain customer trust.
