The Effect of the Expert Concept in AI Agents on Service Attractiveness

The spread of fake news in modern society has rapidly worsened with the advent of generative AI, and is recognized globally as an urgent social problem. While AI agents are attracting attention as a means of information gathering and fact-checking, concerns about hallucinations that provide answers containing false or misleading information necessitate users to verify the sources themselves. In other words, the problem of fake news has yet to be solved. Therefore, the aim of this research is to contribute to solving this problem by limiting the learning scope and usage of AI agents. The research question is: “Does a concept that limits information sources and uses expert knowledge enhance the appeal of AI agents?” Two randomized controlled trials were conducted. STUDY 1 showed that, in the context of AI agent concepts, reliable information curated by experts enhances appeal more than curation or empathy alone. STUDY 2 examined the appeal of the user experience (UX) of expert-based communication in an AI agent used in a clinical setting. The results of this study demonstrated the value of a new perspective, the concept, in the globally competitive AI agent market.