Name: Sidney Scott-Sharoni
Dissertation Proposal Meeting
Date: Monday, June 16th, 2025
Time: 8:30-10:00 AM (Eastern Time)
Mode: Remote (via Teams)

https://teams.microsoft.com/l/meetup-join/19%3ameeting_ZjgwYmNjZmItYzQzZC00ZTk5LWI2N2MtZmQxMjdjNTI5YzBj%40thread.v2/0?context=%7b%22Tid%22%3a%22482198bb-ae7b-4b25-8b7a-6d7f32faa083%22%2c%22Oid%22%3a%221bdbda8b-3d56-4365-a8be-678d28e5e83c%22%7d
    Join conversation
teams.microsoft.com

Dissertation Chairs/Advisor:
Bruce Walker, Ph.D. (Georgia Tech)

Dissertation Committee Members:
Richard Catrambone, Ph.D. (Georgia Tech)
Mengyao Li, Ph.D. (Georgia Tech)
Shadan Sadeghian, Ph.D. (University of Siegen)
Brittany Holthausen, Ph.D. (John Deere)

Title: Challenging the Benefit of Anthropomorphism on Human-AI Collaboration with AI Voice Agents

Abstract: Prevailing theoretical models and frameworks suggest that agents and robots more similar to humans may enhance social relationships, trust, and intention to use; however, recent findings indicate that anthropomorphism may decrease task and collaborative performance. This dissertation tests the assumption that more human-like AI fosters better collaboration by assessing whether a less anthropomorphic and socially intelligent AI voice agent evokes automation bias, leading to stronger conformity and compliance. Across four studies, participants interact with high- and low-social-intelligence voice agents with various manipulations and across four contexts, including trivia, moral decision-making, the Prisoner’s Dilemma task, and in automated vehicle behavior modeling. Initial findings reveal that higher anthropomorphism reduces conformity, despite higher subjective ratings of the more human-like agent, emphasizing a disconnect between preference and behavior. Planned studies will extend this work to trust and compliance. Collectively, the results aim to reevaluate the role of anthropomorphism in social influence and contribute to more effective human-agent collaboration.