The epistemic precondition of democracy is that the public must have an understanding of correct information about the actors involved (Lewandowsky et al., 2017). Democratic systems were built on the assumption that voters disagree over values and opinions rather than disagreement over the correct reality. Citizens must have access to correct information in order to make informed, logical decisions. Nowadays, a significant number of individuals get political information from social media. However, social media also spreads false or misleading content that can confuse voters and shape what they believe (Stachofsky et al., 2023). In this manner, the Internet, which enables citizens to widely access political misinformation, also inhibits their ability to form rational preferences.

Misinformation on social media has a wider and faster reach than true news, (Vosoughi et al., 2018) influencing a large number of individuals and altering their beliefs in candidates, political issues and even swaying their voting decisions. Over time, it can reduce trust in the electoral process. (Mont’Alverne et al., 2024) The effects of misinformation are both attitudinal and institutional in the sense that not only does misinformation affect what people think or believe, but it also weakens democratic legitimacy. Even properly conducted elections can be seen as unfair, an outcome whose consequences include contestation, non-compliance and instability.

Majority of studies focus on public platforms because access to them is unrestricted and easier to research. However, misinformation also spreads on private and semi-private spaces like WhatsApp, Telegram, and direct messages, where people share content with friends and family. These close social ties may lead to people giving more credibility to false claims and challenging them less, which is further augmented by “networked intimacy”. (Rossini et al., 2021).

This paper reviews research on election misinformation on social media and its role in shaping public opinion. It highlights a key gap: we have an incomplete understanding of the ways through which misinformation is circulated differently in public versus private online spaces due to the resistivity of private spaces to fact checking and corrective measures. More specifically, the paper argues that the varying levels of platform visibility drastically alter not only the credibility of information but also the manner in which it is contested.

The discipline that this paper falls under is Political Communication and Digital Media. Political communication encompasses all political discourse, including communication about political actors, which aligns with research on misinformation during elections.(McNair, 2017) The dependent variable (public opinion during elections) has been a central topic of political communication research for over 50 years.(McCombs & Shaw, 1972).

The main argument proposed in this study is that private platforms behave significantly different to public platforms in terms of spreading misinformation in the sense that intimacy and privacy of private messaging changes the response to misinformation received, for example, a person is more susceptible to believe information shared by a relative or a close friend on their private chats than by a stranger publicly posting on the internet. In other words, our realities are governed by a “liquid authority” in which people are more likely to trust information coming from those they know personally rather than from established institutions, therefore changing their perception of what is true. This marks a shift in mindset from institutional authority to relational authority where trust is higher in personal social networks rather than formal verifications.

First, this paper elaborates on the problem in the context of the world and the statistical significance of solving the problem. Secondly, details regarding the literature review are provided, including the main topic, historical discourse on it, contemporary debate around it and the gap through which the research question developed. Thirdly, the paper illustrates arguments and addresses potential counterarguments before concluding.

Various examples of such behaviour are evident in numerous studies that cover real life examples. During the 2016 presidential election (Allcott & Gentzkow, 2017), false stories, such as claims that Hilary Clinton was a criminal conspirator (for example, “Pizzagate”) were propagated on social media, which deepened partisan divides and altered voter perceptions. In the Delhi State Elections (PHILIPOSE, 2015), rumours about candidates and parties were spread on WhatsApp and other private platforms, shaping public opinion in areas where information from trusted contacts is rarely contested.  Similarly, during the Brexit referendum (de Búrca, 2018), the public was tricked into believing claims that suggested misleading information such as the UK sending £350 million a week to the EU, which influenced the public understanding of the financial consequences of leaving the European Union.  The 2017 French presidential election was characterised by “Macron leaks” i.e. hacked documents (some genuine, some altered) released online to change voter perception about Emmanuel Macron. These cases prove the effect that misinformation online has on public opinion in the electoral process.

According to a study done by Vosoughi et al. around 126,000 rumours were spread by upwards of 3 million people on Twitter from 2006 to 2017. These rumours gained wide reach with rapid speed, with the top 1% of false news reaching 1000 to 100,000 people, while true news rarely spread past 100 individuals. (Vosoughi et al., 2018). Other studies have also indicated that this false news and misinformation spread online has a strong influence on voter behaviour and affects the electoral choices they make. (Stachofsky et al., 2023). This difference in the spread of true vs. false information reveals that social media platforms are designed in a way that algorithmically amplifies engaging, sensational content over truth.

Research on this topic is significant as elections are the foundations of a democratic society. Every time an election takes place, there is a plethora of misinformation online about the groups involved. The fake news prevalent online influences voters (Mont’Alverne et al., 2024), leading to them unwittingly giving power to parties that might not be in line with their belief system. (Rhodes, 2024).

On a large scale, this has far-reaching impacts, including a government that does not reflect the aspirations of majority of the population. A long-term implication of the gap between public desire and government actions can be distrust in the electoral process itself. This represents a continuous loop. First, misinformation creates a gap between the objective reality and the public perception, leading to the public voting for candidates who may not align with their political beliefs. When these candidates rise to power, they then do not satisfy the demands of these people, which reinforces distrust. This distrust leads to an increase susceptibility for misinformation and so on. 

Research around the impact of media in shaping public political opinion first appeared in the early 20th century, and later expanded to social media with the onset of features of Web 2.0.

In 1922, Walter Lippmann wrote one of the earliest texts in political communication, putting forward the theory that citizens vote according to “pictures in their heads” created by the media.(Holcombe, 1922). McCombs and Shaw defined the agenda-setting function of media and demonstrated the ways in which the media shapes the way voters view the importance of political issues. (McCombs & Shaw, 1972). Foundational theories such as these establish that the media has an active role in shaping political perception.

From the 1990s to the 2010s, research shifted to digital media. Earlier broadcast models were replaced by networked environments with interactive features and user-generated content. The popular opinion was that more people freely expressing their opinions online would lead to the broadening of people’s horizons and an eventual homogeny in public opinion. Contrary to this, Cass Sunstein introduced the argument that groupism and echo-chambers online let to the fractionation of society and mass polarisation (Nadel, 2002). Manuel Castells studied the rise of scandal politics to show how the media leads to a decline in public trust in democratic systems.(Castells, 2008). Kaplan & Haenlein were the first to define social media platforms as focusing on user-generated content and Web 2.0 features. This was a major step in identifying the factors leading to the spread of misinformation online.(Kaplan & Haenlein, 2010).

In 2016, after the U.S. presidential elections, the term “fake news” became common, and misinformation on social media regarding politics became a popular thematic. In 2017, Lewandowsky et al. proved that a working democracy’s foundation is the public’s understanding of true facts.(Lewandowsky et al., 2017). Miller et al. created the concept of scalable sociability and its implications for political communication.(Miller et al., 2016) This concept is especially relevant in the context of private messaging, which has varying levels that differ from one-on-one direct messaging to larger group chats while remaining out of the public forum.

A considerable number of researchers who have participated in this debate perceive misinformation as a structural problem in social networks. They believe that algorithms create echo chambers that feed us sensational or emotional information that distort our opinions. (Rogers & Niederer, 2020; Yerlikaya & Toker, 2020). This school focuses on the way in which the structure of social media platforms, especially their algorithms, shapes the information that users view on their screens, and how it distorts public understanding.

The other major line of thought holds the user accountable for social media misinformation, and believes that it only shapes public opinion under specific conditions such as weak media literacy or lack of corrective information. (Berger et al., 2025) In this model, individual agency is brought to the forefront and political misinformation is seen as dependent on the user’s susceptibility rather than the structure of the platform.

A common thread between both schools of thought is that they agree that social media impact’s public opinion and affects voter behaviour during elections. The main point of contention between the two schools is whether social media misinformation is impactful by virtue of the structure of social networks or whether it is an outcome of pre-existing social divides. In other words, there is a theoretical tension in political communication literature between individual agency and technological determinism.

The first school of thought fails to acknowledge that users make their own decisions. It solely blames algorithms and platforms for creating an environment in which misinformation thrives. The cases studied here often tend to disproportionately focus on high-stakes election contests where increased levels of tension or uncertainty might cause misinformation to be amplified. This approach could limit the ability to translate their findings in contexts where routine elections are taking place.

The second school, on the other hand, underestimates algorithms that cause repeated exposure to misinformation, over a long period of time. It fails to account for both accounts that propagate and feed polarisation. Media literacy might not be achievable across socio-economic or geographical contexts, and the sheer scale of the corrective information needed may be overwhelming and impossible to manage or be at all feasible.

Both schools of thought have not explored how various group sizes and privacy levels on social media (from direct messaging to group chats to public social media accounts) shape misinformation over time. Private and semi-private spaces are interesting for their impermeability to fact-checking or journalistic correction. Hence, there is a gap in the literature about “visibility gradients” within types of communication on social media. 

Hence, a question on which there is little to no research done arises: How does the circulation of political misinformation across different levels of social media visibility and privacy influence public opinion during electoral campaigns?

The core hunch that this paper was built on is that private messaging environments show distinct misinformation circulation because institutional verification is exchanged for liquid authority.

The assumption that arose while reviewing literature was that private platforms behave significantly different to public platforms in terms of spreading misinformation. This suggests that private messaging changed the response to misinformation received, for example, a person is more susceptible to believe information shared by a relative or a close friend on their private chats than by a stranger publicly posting on the internet.(Rossini et al., 2021) The conditions of trust in the other party and reduced chances of public callouts reduce the incentive for people to fact-check before they share and lead to increased instinctive tendencies (for example, “I personally know and trust the person who gave me this information, therefore it must be credible.”). 

Platforms like WhatsApp could be misused to spread unfounded rumours due to their impermeability to corrective measures due to their private and encrypted nature.(Resende et al., 2019). People are naturally less inclined to challenge information they receive from individuals close to them, especially in one-on-one conversations.(Rossini et al., 2021) Such behaviour can be attributed to the human tendency of social conformity and conflict avoidance i.e. people avoid challenging misinformation in order to avoid personal conflict.

Psychologically speaking, the use of trusted source heuristic and conflict avoidance norms decreases the likelihood of corrective behaviour, while sociologically, Giddens’ change from systemic trust to personal trust offers insight into why relational authority takes precedence over institutional authority (Giddens, 1990). The psychological theory of dual process model by Kahneman sheds light on why reduced effort in sharing misinformation leads people becoming more vulnerable to false claims (Krämer, 2014).Another relevant psychological theory is Festinger’s Social Comparison Theory and Cognitive Dissonance Theory. (Festinger, 1954)

People generally place their trust in news agencies as opposed to social media, but their response to online misinformation is greatly altered through private messaging and sharing, when information is received from close social ties such as friends or family.

Empirical studies (Garimella & Eckles, 2020; Rossini et al., 2021) show that misinformation shared within close personal networks is retained more and corrected less than information on public platforms. It was found that nearly a quarter of the group in a longitudinal study reported sharing misinformation on WhatsApp or Facebook. Behaviours that influence political outcomes, such as discussing politics, utilising social media as a source of news, and “being exposed to cross-cutting opinions” are also associated with dysfunctional information sharing. (Rossini et al., 2021) Studies conducted in rural India indicate that majority of the misinformation shared through private messaging is political in nature and influences political narratives and electoral choices.(Garimella & Eckles, 2020) These studies suggest that private platforms are key channels of political misinformation, and that public opinion is greatly shaped by the information received in closed communication spaces rather than on public platforms.

The closed nature of private messaging and ease with which it is possible to share information on a large-scale to a variety of groups and individuals make private messaging platforms extremely resistive to measures that prevent or correct misinformation. “WhatsApp opens a paradoxical use of its platform, allowing at the same time the viral spread of a content and encrypted personal chat.”(Melo et al., 2019). Such platforms enable forwarding and propagation of information on a wide scale. Certain limitations and restrictions imposed by these networks are cannot effectively counter misinformation. “A content can spread quite fast through the network structure of public groups in WhatsApp, reaching later the private groups and individual users.” A study conducted on political WhatsApp discovered that the content propagated was largely misinformation.(Garimella et al., 2025).

Platform architecture is important: Algorithmic ranking, ease of forwarding, restrictions on group size, and engagement motivations are structural elements that help misinformation thrive. The responsibility for such elements, then, lies also with the developers and policymakers of platform architecture (Gillespie, 2018).

Generative artificial intelligence has changed the information ecosystem both with regards to the production as well as the validation of information. Earlier forms of misinformation required human effort to create and disseminate. However, GenAI has enabled the rapid spread of misinformation at low costs. This misinformation also tends to be more persuasive, scalable and context specific than human generated misinformation. (Goldstein et al., 2023). As a result, malicious actors can now change political narratives across both public and private networks with more ease than they have ever had before, causing an unprecedented rise in misinformation in terms of volume and speed.

More importantly, GenAI systems draw information from large databases taken off the internet, which include both verified information and widely circulated misinformation. (Bommasani et al., 2021; Brown et al., 2020). When misinformation is repeated many times online, it can become embedded in training datasets used for machine learning. As a result, the software presents these falsehoods as verified claims when asked about them, making misinformation seemingly legitimate.(Bender et al., 2021).

Therefore, a continuous feedback loop is generated in which misinformation travels across social media networks, is propagated widely and eventually is incorporated into GenAI training databases. This data is further reproduced by AI systems who feed it back into human social media networks. In private networks, AI generated misinformation face even lower levels of scrutiny due to the combined effect of trust in the sender and the context-specific content. (Goldstein et al., 2023).

In the context of political communication, there is a clearly visible shift from platform mediated misinformation to a more hybrid form of misinformation where epistemic authority lies both in human relationships as well as computational outputs. Hence, not only is it easier to produce and spread misinformation, but the credibility of this information is enhanced through machine learning systems that cannot distinguish between widely believed falsehood and verified truths. (Bender et al., 2021; Bommasani et al., 2021).

Some scholars argue that users actively try to correct information in private networks rather than spread it (Vijaykumar et al., 2021), however most users either ignore misinformation and do not actively intervene, and the rate of successful corrections is very low compared to the rate at which false information spreads. (Melo et al., 2019) Corrective intent may exist, but the empirical evidence indicates that the effectives of any measures taken is hindered by privacy levels and low platform visibility, the very features that facilitate the spread and increased credibility of misinformation. 

The research so far clearly shows that social media influences people’s beliefs and priorities during elections, and that misinformation spreads at a rapid rate to a large group. Many studies also agree that misinformation can affect voter opinions and sometimes even electoral decisions. (Rossini et al., 2021).

Despite this, a previously unaddressed area remains unexplored in this field. While recent studies have started looking at misinformation on private and semi-private platforms like WhatsApp, research remains scattered and disconnected. While some focus on the nature of misinformation propagation, others turn their attention to a few specific case studied or surveys. (Resende et al., 2019) However, there is a lack of a clear framework that compares public platforms and private messaging spaces systematically, especially across different levels of visibility and privacy.

The reason for this gap is partly due to practical research problems such as encrypted messaging platforms, which make studying misinformation in private spaces difficult on a large scale. Due to this, researchers have had to use indirect methods such as surveys, small samples or data donation project which, while useful, give rise to issues on effective data recording and analysis.

The argument presented in this paper is limited by a lack of empirical data due to the encrypted nature of private platforms. Large-scale misinformation spread is not easily studied and thus most existing studies rely on surveys or data samples which are incapable of capturing real-world behaviour. Also, the paper risks ignoring the influence of public platforms. In reality, misinformation does not exist solely in the private or the public space but travels between the two. For example, a false fact might appear on Instagram, but it travels to the private space when a user forwards the link to a close friend on WhatsApp. Thus, solely focusing on private platforms fails to fully capture the manner in which misinformation influences voter behaviour. Future research based on the intersection of misinformation flow across both the private and public platforms is needed in order to fully understand exactly how misinformation on social media shapes public opinion.

As a result, it remains unclear how misinformation works differently across public and private online spaces during elections. We do not yet have answers to questions like how trust in close relationships (like friends and family) changes how people respond to misinformation. This remains an open and important area for future research in political communication and digital media.

1. Allcott, H., & Gentzkow, M. (2017). Social Media and Fake News in the 2016 Election. The Journal of Economic Perspectives, 31(2), 211–235. JSTOR.
2. Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021). On the Dangers of Stochastic Parrots: Can Language Models Be Too Big? 🦜. Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, FAccT ’21, 610–623. https://doi.org/10.1145/3442188.3445922
3. Berger, L. M., Kerkhof, A., Mindl, F., & Münster, J. (2025). Debunking “fake news” on social media: Immediate and short-term effects of fact-checking and media literacy interventions. Journal of Public Economics, 245, 105345. https://doi.org/10.1016/j.jpubeco.2025.105345
4. Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., Arx, S. von, Bernstein, M. S., Bohg, J., Bosselut, A., Brunskill, E., Brynjolfsson, E., Buch, S., Card, D., Castellon, R., Chatterji, N. S., Chen, A. S., Creel, K. A., Davis, J., Demszky, D., … Liang, P. (2021). On the Opportunities and Risks of Foundation Models. ArXiv, abs/2108.07258. https://api.semanticscholar.org/CorpusID:237091588
5. Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., Agarwal, S., Herbert-Voss, A., Krueger, G., Henighan, T., Child, R., Ramesh, A., Ziegler, D. M., Wu, J., Winter, C., … Amodei, D. (2020). Language Models are Few-Shot Learners. arXiv. https://doi.org/10.48550/arXiv.2005.14165
6. Castells, M. (2008). The New Public Sphere: Global Civil Society, Communication Networks, and Global Governance. The Annals of the American Academy of Political and Social Science, 616, 78–93. JSTOR.
7. de Búrca, G. (2018). How British was the Brexit vote? In B. Martill & U. Staiger (Eds), Brexit and Beyond (pp. 46–52). UCL Press. JSTOR. https://doi.org/10.2307/j.ctt20krxf8.10
8. Festinger, L. (1954). A Theory of Social Comparison Processes. Human Relations, 7, 117–140.
9. Garimella, K., Cintaqia, P., Constain, J. J. R., Nayak, B., & Vashistha, A. (2025). Global Patterns of Viral Content on WhatsApp (arXiv:2407.08172). arXiv. https://doi.org/10.48550/arXiv.2407.08172
10. Garimella, K., & Eckles, D. (2020). Images and Misinformation in Political Groups: Evidence from WhatsApp in India (arXiv:2005.09784). arXiv. https://doi.org/10.48550/arXiv.2005.09784
11. Giddens, A. (1990). The Consequences of Modernity. Stanford University Press. https://books.google.co.in/books?id=C46N9wtBI0gC
12. Gillespie, T. (2018). Custodians of the Internet: Platforms, Content Moderation, and the Hidden Decisions that Shape Social Media. Yale University Press. https://books.google.co.in/books?id=-RteDwAAQBAJ
13. Goldstein, J. A., Sastry, G., Musser, M., DiResta, R., Gentzel, M., & Sedova, K. (2023). Generative Language Models and Automated Influence Operations: Emerging Threats and Potential Mitigations. ArXiv, abs/2301.04246. https://api.semanticscholar.org/CorpusID:255595557
14. Holcombe, A. N. (1922). Public Opinion. By Walter Lippmann. (New York: Harcourt, Brace and Company. 1922. Pp. X, 427.). American Political Science Review, 16(3), 500–501. Cambridge Core. https://doi.org/10.2307/1943740
15. Kaplan, A. M., & Haenlein, M. (2010). Users of the world, unite! The challenges and opportunities of Social Media. Business Horizons, 53(1), 59–68. https://doi.org/10.1016/j.bushor.2009.09.003
16. Krämer, W. (2014). Kahneman, D. (2011): Thinking, Fast and Slow. Statistical Papers, 55(3), 915–915. https://doi.org/10.1007/s00362-013-0533-y
17. Lewandowsky, S., Ecker, U. K. H., & Cook, J. (2017). Beyond misinformation:Understanding and coping with the ‘post-truth’ era. Journal of Applied Research in Memory and Cognition, 6(4), 353–369.
18. McCombs, M. E., & Shaw, D. L. (1972). The Agenda-Setting Function of Mass Media. The Public Opinion Quarterly, 36(2), 176–187.
19. McNair, B. (2017). An Introduction to Political Communication: Sixth Edition (Sixth). https://www.perlego.com/ereader/1561330
20. Melo, P. de F., Vieira, C. C., Garimella, K., Melo, P. O. S. V. de, & Benevenuto, F. (2019). Can WhatsApp Counter Misinformation by Limiting Message Forwarding? (arXiv:1909.08740). arXiv. https://doi.org/10.48550/arXiv.1909.08740
21. Miller, D., Costa, E., Haynes, N., McDonald, T., Nicolescu, R., Sinanan, J., Spyer, J., Venkatraman, S., & Wang, X. (2016). Academic studies of social media. In How the World Changed Social Media (1st edn, Vol. 1, pp. 9–24). UCL Press. https://doi.org/10.2307/j.ctt1g69z35.9
22. Mont’Alverne, C., Arguedas, A. R., Banerjee, S., Toff, B., Fletcher, R., & Nielsen, R. K. (2024). The Electoral Misinformation Nexus: How News Consumption, Platform Use, and Trust in News Influence Belief in Electoral Misinformation. Public Opinion Quarterly, 88(SI), 681–707. https://doi.org/10.1093/poq/nfae019
23. Nadel, M. S. (2002). Customized News Services and Extremist Enclaves in ‘Republic.com’. Stanford Law Review, 54(4), 831–886. JSTOR. https://doi.org/10.2307/1229580
24. PHILIPOSE, P. (2015). Can an Election Be Tweeted to Victory? AAP, Arvind and a City State Called Delhi. Economic and Political Weekly, 50(8), 10–13. JSTOR.
25. Resende, G., Melo, P., Sousa, H., Messias, J., Vasconcelos, M., Almeida, J., & Benevenuto, F. (2019). (Mis)Information Dissemination in WhatsApp: Gathering, Analyzing and Countermeasures. The World Wide Web Conference, 818–828. https://doi.org/10.1145/3308558.3313688
26. Rhodes, S. (2024). Narrow Margins and Misinformation: The Impact of Sharing Fake News in Close Contests. Social Sciences, 13(11), 571. https://doi.org/10.3390/socsci13110571
27. Rogers, R., & Niederer, S. (2020). The politics of social media manipulation. In R. Rogers & S. Niederer (Eds), The Politics of Social Media Manipulation (pp. 19–70). Amsterdam University Press. JSTOR. https://doi.org/10.2307/j.ctv1b0fvs5.3
28. Rossini, P., Stromer-Galley, J., Baptista, E. A., & Veiga De Oliveira, V. (2021). Dysfunctional information sharing on WhatsApp and Facebook: The role of political talk, cross-cutting exposure and social corrections. New Media & Society, 23(8), 2430–2451. https://doi.org/10.1177/1461444820928059
29. Stachofsky, J., Schaupp, L. C., & Crossler, R. E. (2023). Measuring the effect of political alignment, platforms, and fake news consumption on voter concern for election processes. Government Information Quarterly, 40(3), 101810. https://doi.org/10.1016/j.giq.2023.101810
30. Vijaykumar, S., Rogerson, D. T., Jin, Y., & De Oliveira Costa, M. S. (2021). Dynamics of social corrections to peers sharing COVID-19 misinformation on WhatsApp in Brazil. Journal of the American Medical Informatics Association, 29(1), 33–42. https://doi.org/10.1093/jamia/ocab219
31. Vosoughi, S., Roy, D., & Aral, S. (2018). The spread of true and false news online. Science, 359(6380), 1146–1151. https://doi.org/10.1126/science.aap9559
32. Yerlikaya, T., & Toker, S. (2020). Social Media and Fake News in the Post-Truth Era: The Manipulation of Politics in the Election Process. Insight Turkey, 177–196. https://doi.org/10.25253/99.2020222.11