Generative AI Source Credibility in Fact-Checking: The Role of Layered Sources
Year: 2026
Type: article
Abstract: This study examines how layered AI source cues shape credibility perceptions (expertise, trustworthiness, and goodwill) and belief change in AI fact-checking. AI as a source was conceptualized through two layered source cues: sponsorship (outer-layer cue) and training data (inner-layer cue). A 3 (sponsorship: government–CDC, nonprofit–Mayo Clinic, or for-profit–Pfizer) × 2 (training data: internet data or medical-journal fine-tuning) online experiment in vaccine misinformation fact-checking found no significant effects of sponsorship on any credibility dimensions, while training data fine-tuning increased perceived expertise. Neither cue indirectly affected belief correction through credibility. Findings clarify AI source cues’ value in fact-checking.
Authors: Joana Videira, Jingbo Meng, Bingjie Liu, Wei Peng
Institutions: The Ohio State University, Michigan State University
Language: English
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W7233596927
