The Effect of Artificial Intelligent (AI) Agent’s Self-Disclosures on Human-AI Interaction Outcomes

With increasing involvement Artificial Intelligenct (AI) technology in our daily lives, we need to learn about factors that can facilitate people’s trust in and liking of an AI, which can promote better human-AI interactions and eventually help humans to benefit from the service that AI provides. Self-disclosing behaviours, or acts of revealing one’s personal information to another, are crucial to interpersonal relationship development among people. We have found in our previous experiments that people tend to like those who appear to be more similar to themselves through self-disclosures. In this project, we will test the effectiveness of using AI’s self-disclosing behaviours to facilitate liking and trust from its human partner. Specifically, using the virtual human software developed by Dr. Gratch’s lab, we will investigate whether people will like an AI more when it reveals greater similarity to themselves through self-disclosures. TO BE CONT’D

Faculty Supervisor:

Erin Heerey

Student:

Partner:

University of Southern California

Discipline:

Sociology

Sector:

Education

University:

Western University

Program: