AI-driven advisory in insurance services : An experimental study on trust, transparency, and policy adoption (Journal of Information and Optimization Sciences)

Summary

• The study evaluates AI-only, human-only, and hybrid human-AI advisory models through a randomized experiment involving 90 insurance consumers.p.2
• Participants were evenly assigned to three experimental conditions, with 30 consumers placed in each advisory model group.p.6

Key insights

1. Hybrid human-AI advisory generates significantly higher consumer trust, perceived transparency, and policy purchase intent than pure human or algorithmic models.p.2
2. Deploying an AI-only advisory service improves efficiency but heightens cyber security and privacy concerns, which impedes consumer adoption.p.2
3. Higher AI literacy positively moderates user acceptance and trust in AI-only advisory services with an interaction coefficient of β = .27.p.8
4. Consumers with negative past insurance claim experiences demonstrate stronger trust and purchase preferences toward human or hybrid advisory modes.p.8
5. Cyber security concerns significantly weaken the positive linkage between AI-only advisory and purchase intention with β = −.31.p.8