A recent study by researchers from Taiwan has highlighted the powerful role of political bots in amplifying societal divisions on social media. According to the study, these automated accounts, designed to spread specific political agendas, are significantly contributing to the polarisation of public opinion online.
The study, published in Social Science Computer Review, explores how political bots, through their strategic placement within social networks, are able to drive people towards extreme viewpoints, thereby deepening divisions within society. Using an agent-based model to simulate social networks, the researchers demonstrated that the presence of bots, particularly when deployed symmetrically on both sides of the political spectrum, intensifies opinion polarisation.
Political bots, which are algorithmically programmed to achieve political objectives, have become a pervasive force within online social networks such as Twitter and Facebook. These bots can mimic human behaviour by posting, commenting, and sharing content, making them difficult to detect. Their activity has been linked to the spread of misinformation and propaganda, further complicating the online discourse.
Hsuan-Wei Lee, PhD, assistant professor at National Taiwan University and co-author of the study, said: “The study was motivated by the growing concern over the role of political bots in shaping public opinion and exacerbating polarisation within social networks. As social media platforms have become a primary source of information and political discourse, the presence of bots—automated accounts designed to manipulate discussions – has raised alarms about their ability to influence public sentiment.”
The study focused on how these bots interact within complex social networks. By introducing bots into a model of bounded-confidence opinion dynamics, they observed that bots, regardless of their number, could cause significant polarisation, especially when they occupy central positions within the network. The simulations revealed that even a small number of bots strategically placed in highly connected parts of the network could have a disproportionate impact on public opinion.
“The study’s simulations reveal that the symmetrical deployment of political bots across both sides of the opinion spectrum significantly intensifies polarisation within social networks,” Lee explained. “Polarisation is particularly pronounced when individuals have moderate levels of tolerance and when bots are strategically placed in networks with short average path lengths.”
The researchers found that the polarising effect of bots is most pronounced when they are symmetrically placed on both ends of the ideological spectrum. This symmetrical deployment leads to a situation where opposing viewpoints are reinforced, pushing users towards more extreme positions. The study highlighted that this effect is contingent on certain network characteristics, such as the average path length and the degree of homophily – where users are more likely to interact with others who share similar views.
Interestingly, the study also discovered that the impact of bots varies depending on the structure of the network. In networks with high levels of clustering, where users are closely connected to a small group of similar individuals, the polarising effect of bots was less severe. However, in networks where information could spread quickly across a broader range of users, polarisation increased more rapidly.
Another key finding was the role of user tolerance in this process. The researchers noted that polarisation tends to decrease in environments where users have either very low or very high tolerance for differing opinions. But in situations where users are moderately tolerant, bots are more likely to push them towards more extreme viewpoints, leading to greater polarisation.
Lee highlighted the implications of these findings: “The findings suggest that bots not only push public opinion towards extremes but also accelerate the process of polarisation when strategically embedded within key network positions. This has important implications for understanding the dynamics of digital influence and underscores the need for strategies to mitigate the polarising effects of bots.”
The study offers important insights into the ongoing debate about the role of social media in polarising public opinion. It suggests that the presence of bots is not just a minor nuisance but a significant driver of division within society. The findings are particularly relevant in the context of major political events, such as elections, where the manipulation of public opinion through social media can have far-reaching consequences.
The study also raises questions about the effectiveness of current efforts to combat bot-driven polarisation. While platforms like Twitter and Facebook have implemented measures to identify and remove bots, the study suggests that more needs to be done to address the underlying dynamics that allow bots to exert such a powerful influence.
“The study suggests that interventions aimed at reducing the centrality of bots in networks and increasing public awareness of bot-driven manipulation could be effective in curbing their impact,” Lee concluded.
Future research will incorporate real-world data from platforms like X and Facebook to validate and refine the findings of this study. By integrating empirical data with the existing model, the research aims to provide a more comprehensive understanding of the role of bots across different social media environments. Additionally, future work will explore the impact of platform-specific features, such as content moderation policies and network structures, on the behaviour and influence of bots, offering practical insights for mitigating polarisation in digital spaces.









