Lindsay Louwers
Research Mentor: Araceli Cruz
Mentor Department: Political Science, LSA
Author(s): Cesi Cruz
Session: Session 4 (1:00 PM – 1:50 PM)
Presentation Type: Poster 49
Abstract
This project is developing a reflective AI chatbot designed to build resilience against misinformation and polarization. Unlike traditional fact-checking or content moderation approaches, the chatbot does not issue direct corrections. A different approach is taken; this chatbot uses large language models (LLMs) to foster reflection along with an identity-aware dialogue. It prompts users to consider the why behind the information they are receiving. As social media platforms amplify emotionally charged narratives, individuals may become more entrenched in identity-based viewpoints rather than open to correction, especially from AI. This chatbot is designed with an awareness that beliefs can be shaped by more than facts: they’re also shaped by who we are, what we’ve experienced, and what communities we belong to.
