Processing Fabricated Foreign Risks with AI Fact-Checker

The role of Third-Person Perception, Nationalism and Conspiratorial Orientation

Abstract

The proliferation of fabricated information presents a global challenge, particularly when infused with conspiratorial narratives. As AI technologies become integrated into news production and verification, interfaces increasingly display labels such as “verified by a human” or “verified by AI,” altering how individuals evaluate risk-related content. Personality traits—especially blind nationalism and conspiratorial orientation—shape responses to such information, while perceptions of social norms and susceptibility (third-person perception, TPP) mediate these effects. This study investigates how verification source (human vs. AI) and personality traits interact to influence third-person perception and subsequent information-seeking intentions. In a between-subjects experiment (N = 506), participants were randomly assigned to read misinformation labeled as verified either by a human or by AI, then completed measures assessing perceived social norms, third-person perception (TPP), and information-seeking intentions. Results show that H1 was supported: blind nationalism consistently predicted higher perceived social norms of the foreign-risk misinformation in both conditions (human: β = 0.465, p < .001; AI: β = 0.371, p < .001). H3 was supported: higher perceived norms were associated with lower third-person perception in both conditions (human: β = –0.244, p = .003; AI: β = –0.364, p < .001). H4 was supported: lower TPP predicted stronger intentions to seek additional information (human: β = –0.211, p < .001; AI: β = –0.208, p < .001). The indirect pathway from nationalism to information-seeking via perceived norms and TPP was significant in both groups, with comparable indirect effects across conditions. In contrast, H2 was not supported: conspiratorial orientation had no significant effect on perceived norms or on any downstream outcome in either condition. These findings underscore that verification interface moderates the psychological pathways through which personal predispositions shape engagement with misinformation. The third-person effect emerges as a key mediator, translating normative perceptions into behavioral intentions. This study contributes to cross-contextual understanding of AI-mediated news credibility and offers practical insights for designing verification systems that enhance public resilience to misinformation.