Platform-generated misinformation warning labels during disasters: A U.S.-China comparative study

This study introduces a platform social responsibility concept and a platform-public relationship perspective, identifying two crucial warning label features—source agent and communication style—to explore their impact on perceived misinformation credibility, platform credibility, platform social responsibility, and platform support behavior. Using a 2 source agents (AI vs. human) x 2 communication styles (direct vs. indirect) between-subjects online experiment conducted in the U.S. (US) and China, the study finds that warning labels generated by human fact-checkers are more effective than those from AI fact-checkers in reducing misinformation credibility and enhancing platform credibility, social responsibility, and platform support behavior. The study also reveals distinct mechanisms underlying warning label effectiveness in the US and China. In the US, disaster responsibility attribution functions as a boundary condition, influencing how warning labels function as veracity signaling, primarily affecting misinformation credibility. In contrast, in China, AI heuristics serve as a boundary condition, extending the impact of warning labels beyond misinformation credibility to strengthen platform-public relationships through enhancing platform credibility, social responsibility, and platform support behavior. These findings provide new insights into the role of warning labels in combating misinformation, highlighting the importance of cultural differences and the need for tailored platform interventions.