Investigating customer trust across human-AI collaboration styles

As AI becomes a collaborative partner in professional decision-making, understanding how human–AI collaboration styles shape customer trust is increasingly critical. This research investigates how two collaboration styles, ongoing versus deferred, affect customer trust. Drawing on theories of human-AI collaboration and delegation, we propose that ongoing collaboration, where a human advisor actively engages with AI throughout the process, fosters greater trust than deferred collaboration, where the advisor intervenes only after the AI has completed its analysis. Two experimental studies tested this proposition in financial and well-being service contexts. Results show that ongoing collaboration significantly increases trust in both the recommendation and the firm. This effect is sequentially mediated by increased perceptions of elaboration and infusion. The findings contribute to information system literature by advancing a dyadic view of human–AI teaming, emphasizing the importance of responsibility sharing. Practically, the research offers guidance for designing AI-assisted services that emphasizing ongoing human–AI collaboration can enhance customer confidence. These insights underscore the importance of how, not just whether, AI is integrated into service delivery.