October 27, 2025

Misinformation Detection by AI Chatbots and Humans

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Comparative analysis reveals AI chatbots improve misinformation detection over humans, suggesting biases exist in both groups.

AI chatbots detect misinformation more accurately than U.S. Democrats and Republicans, showing high effectiveness.
In two preregistered studies with 2,461 participants, discrepancies between AI and human performance were significant.
Both AI and humans displayed biases in accepting or rejecting information regardless of veracity, indicating common pitfalls.
Findings challenge misconceptions of AI as purely misleading entities while highlighting their nuanced biases.

Nahon et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark