Unveiling the impact of network structure on misinformation co-sharing among social bots and human users

Purpose The literature has revealed various factors driving misinformation sharing in the online sphere. However, there has been a lack of research on the structural factors that impact the spread of misinformation, and the specific roles played by social bots in this context remain largely unexplored. Design/methodology/approach Using a social network approach, we analyzed a backbone network of co-sharing among Twitter (now X) accounts connected by sharing the same links from low-credibility news websites, constructed from 146,580 tweets. Findings Our findings reveal that accounts with varying bot likelihoods are more likely to connect, supporting bot-based heterogeneity over homophily. The co-sharing network was also affected by structural factors, such as the tendency for popular accounts to attract more connections (known as preferential attachment), and individual account characteristics. The “echo chamber” hypothesis was not supported in the context of misinformation co-sharing. Originality/value These results highlight the significant impact of social bots in spreading misinformation. Peer review The peer review history for this article is available at: