Probing Multimodal Large Language Models on Cognitive Biases in Chinese Short-Video Misinformation
For researchers and developers of MLLMs, this work highlights vulnerabilities to cognitive biases in multimodal misinformation, though the dataset is limited to Chinese short videos and four health domains.
The paper introduces a framework to evaluate Multimodal Large Language Models (MLLMs) on cognitive biases in Chinese short-video misinformation, using a dataset of 200 videos with fine-grained annotations. Gemini-2.5-Pro achieves the highest belief score of 71.5/100, while o3 scores 35.2, and models are susceptible to biases like authoritative channel IDs.
Short-video platforms have become major channels for misinformation, where deceptive claims frequently leverage visual experiments and social cues. While Multimodal Large Language Models (MLLMs) have demonstrated impressive reasoning capabilities, their robustness against misinformation entangled with cognitive biases remains under-explored. In this paper, we introduce a comprehensive evaluation framework using a high-quality, manually annotated dataset of 200 short videos spanning four health domains. This dataset provides fine-grained annotations for three deceptive patterns-experimental errors, logical fallacies, and fabricated claims-each verified by evidence such as national standards and academic literature. We evaluate eight frontier MLLMs across five modality settings. Experimental results demonstrate that Gemini-2.5-Pro achieves the highest performance in the multimodal setting with a belief score of 71.5/100, while o3 performs the worst at 35.2. Furthermore, we investigate social cues that induce false beliefs in videos and find that models are susceptible to biases like authoritative channel IDs.