Abstract

Amid the COVID-19 pandemic, a parallel crisis emerged in the form of an “infodemic,” where misinformation proliferated through social media at an unprecedented scale. This article delves into the evolving landscape of reliable and misinformation sources reporting on the pandemic, examining their communication dynamics. Leveraging a comprehensive data set encompassing 437,832 news published online by media organizations from January 2020 to December 2021, I employ structural change analysis and network-based natural language processing to explore these dynamics. The findings illuminate the contrasting approaches adopted by reliable sources, which prioritize scientific discourse, and misinformation, which exhibits high topic volatility—rapid change in the focus—suggesting that misinformation sources frequently shift their attention to different topics. Notably, this highlights a deliberate strategy employed by misinformation purveyors, that is, creating false crises without authoritative resolutions, thereby fostering a sense of lack of control, chaos, and uncertainty. Furthermore, the sentiment and morality analysis of COVID-19 news reveal that reliable news tends to maintain a balanced and neutral tone, while misinformation often exhibits a highly negative and morally charged narrative. These results carry significant implications for devising effective social countermeasures against misinformation at various stages of the pandemic.