Chat but Verify: Combating Misinformation by Generative AI with Verification Affordance,

Mengqi (Maggie) Liao and Sundar, S.S. “Chat but Verify: Combating Misinformation by Generative AI with Verification Affordance,” paper to be presented at the 75th annual conference of the International Communication Association, Denver, June.

Abstract: Conversationality is the hallmark of Generative AI tools like ChatGPT. But could it be the reason why lay users readily trust AI-generated information? If so, it can be a problem because these tools are known to hallucinate and even generate misinformation. Will adding a verification mechanism help mitigate this problem? Guided by the HAII TIME model, we addressed these questions with a pre-registered 2 (Gen-AI conversationality: low vs. high) x 4 (verification affordance: absence vs. cue vs. forced action vs. voluntary action) between-subjects online experiment measuring responses to a Gen-AI agent that disseminates misinformation (N=477). We found that high conversationality potentiates the social presence heuristic and engenders higher trust. However, acting on the verification affordance mitigates this problem by decreasing perceived credibility of the generated information, and inhibiting positive machine heuristic and Magical AI heuristic. Theoretical and practical implications are discussed regarding human-AI interaction and designing more socially responsible technology.