Truth, Trust, and Technology: Political Communication in the Era of Synthetic Media

Truth, Trust, and Technology: Political Communication in the Era of Synthetic Media | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Truth, Trust, and Technology: Political Communication in the Era of Synthetic Media Nigar Garajamirli This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9473319/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 3 You are reading this latest preprint version Abstract In this study computational social science methods are implemented to large-scale election datasets. The present study aims to examine the structural dynamics of the political communication within the landscape of social media. At the same time it conducts a comparative analysis of patterns related to communication in the 2016 and 2024 US presidential elections, which focus on participation disparities, message amplification, automated behavior, and network architecture. Alongside with the study’s framework, statistical analysis, network modeling and MATLAB simulations are used aiming at examining the ways of how political information spreads and gains visibility in social media realm. The results can be interpreted as the ones with clear imbalances. It should be noted, that only a tiny group of users produces most of the content seen in social media, while others tend to remain less active. Message usage can be seen as not even and only a few messages could in result achieve wide dissemination. The findings also indicate that bot-like behavioral characteristics, which indicate that these bots are inclined to produce and disseminate content at a much higher rate comapring to the average user. A notable pattern emerges indicating that actors with strong connections act in the roles of central hubs and highly contribute to shaping the dissemination of the information. The comparative analysis showed a significant relationship between that dissemination power, network centralization, and the reach of active accounts, which increased in 2024. It is worth noting that the results of the current simulation indicated that even with a small number of automated accounts, they can significantly increase information reach. Humanities/Complex networks Social science/Complex networks Physical sciences/Mathematics and computing Physical sciences/Physics Digital political communication computational social science information diffusion algorithmic amplification automated accounts network centralization Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 1. Introduction The fast spread of the digital communication technologies has led to certain radical changes in the ways how we communicate politically and tend to participate in democratic life. Today, social media platforms play a role of the central forums for political debatштп, which enable politicians, journalists, activists, and citizens to be able to interact within a scope of highly interconnected media environment. These platforms can facilitate the fast enough dissemination of information, real-time political engagement and the mobilization of public opinion on a pretty large scale, which fundamentally alters the means how political discourse is shaped and in result disseminated in modern society (Tucker et al., 2018 ; Kreiss, 2016 ; Jungherr, 2016 ; Schillemans, 2014 ). Notably in this digital ecosystem, communications are seen to be increasingly mediated by algorithmic infrastructure, which in turn organizes, categorizes, and distributes political content across the entire network (Napoli, 2019 ; Gillespie, 2018 ; Helberger et al., 2018 ; Bucher, 2018 ; Beer, 2017 ). A key point is that, with the increasing influence of the infrastructure of platforms and algorithmic systems new opportunities are notably emerging in the realms of political movements and civic engagement. As a consequence, politicians are believed to increasingly rely on strategies with the data being in the core, personalized messaging and targeted advertising to be able to reach voters online (Aral & Eckles, 2019 ; Buts, 2021; Kreiss et al., 2018 ). As a result, digital platforms tend to become deeply integrated into the process of the modern election, which in turn lead to fundamental changes in the strategies of the traditional election and expanding the role of online debate regarding the political discourse altogether shaping public opinion (Karpf, 2024; Bimber, 2014 ). On the other hand, the technological infrastructure that makes possible broad political participation tend to create new vulnerabilities in the communication systems of the democratic world. A plenty of studies indicate that the fast enough spread of misinformation, the campaigns of the organized disinformation alongside with the emergence of artificial media may pose truly big challenges to modern democracy (Lazer et al., 2018 ; Wardle & Derakhshan, 2017 ; Farkas & Schou, 2020; Benkler et al., 2018 ). The misinformation and falsehoods are believed to be spread rapidly on social media, which in turn often reaching many people, sometimes even faster than the verified information (Vosoughi et al., 2018 ). Therefore, the exposure to content of this type can be often limited to specifically defined ideological communities, resulting in reinforcing the existing belief systems among the people and exacerbate the polarization among politics (Guess et al., 2018 ; Grinberg et al., 2019 ; Cinelli et al., 2020). It is worth noting that, the recent up-to-date advances in the area of the artificial intelligence and generative technologies have made these challenges worse. As a consequence, AI-generated texts, images, and videos (including "deepfakes") tend to create the new forms of artificial communication in politics which is capable of manipulating political discourse and influencing public opinion (Battista, 2024; Kohanets, 2025; Bommasani et al., 2021 ; Floridi et al., 2018 ). Therefore, these technologies make the automated production of massive amounts of interactive content possible, which in result make it pretty difficult for users to find the difference between the original and automated information. In some possible cases, the engagement with AI-generated content was close to equal or even it was possible to surpass engagement with the content generated with human, which could raise concerns about the potential for possible manipulation of the environment of digital information. One of the most well-known and significant mechanisms in terms of artificial influence in online political communication is the utilization of of automated social media accounts, which are widely known as "social bots." These bots are believed to be algorithmically controlled accounts, which are capable of generating content highly independently, interacting with other users alongside with other functions like amplifying political messages within digital networks (Ferrara et al., 2016 ; Howard & Kollanyi, 2016; Ferrara, 2017 ). In accordance with the experimental studies, organized botnets can somehow be the reason of distorting political discourse via generating massive amounts of messages and at the same time artificially amplify certain narratives (Bessi & Ferrara, 2016 ; Shao et al., 2018 ; Subrahmanian et al., 2016 ). Moreover, organized bot activity is tended to be observed during events regarding political activities like national elections, where it strongly manipulates the flow of information and could shape online discussions via using the strategies commonly recognized as computational propaganda (Badawy et al., 2018 ; Bradshaw & Howard, 2018; Woolley & Howard, 2017 ). The possible presence of automated actors within social media platforms can be considered as a significant threat to the communication within democratic society. These automated programs can amplify unreliable sources, reinforce closed ideological affiliations, and spread misleading narratives within divided societies (Cinelli et al., 2020; Bovet & Makse, 2019 ; Stella et al., 2018 ). During election campaigns, organized disinformation activities are tend to be increasingly documented, taking into account the automated accounts, which strategically spread political claims and misinformation in order to influence public opinion and behavior of the voter (Cinus et al., 2024 ; Im et al., 2020). Despite the fact, that the research regarding the disinformation via computer generation and propaganda could make great strides, a significant gap still remains nowadays in the empirical understanding in terms of the dynamics of artificial political communication (Das & Clark, 2019 ). To be precise, a deeper understanding of the ways how automated actors tend to influence political discourse in the environments of the social media requires the integrated approach based on computations that can simultaneously analyse the patterns of the engagement, network structures altogether with amplification dynamics. Even though previous research had a focus of analysing bot detection and disinformation dissemination separately, relatively few studies are believed to integrate these analytical perspectives into a unified framework of the computations (Ferrara, 2017 ; Shao et al., 2018 ; Cresci et al., 2017 ). With the aim of the bridging this gap, this study conducts a series of computational analyses in order to examine patterns of communication within the political environment and artificial amplification within social networks. Via extensive Twitter data on election-related discussions, author analyses the distribution of all user activity, the dynamics of retweet amplification, metrics of bot-like behavior, finishing with the the network architecture that can potentially shape information dissemination. Furthermore, author could employ modelling within statistical frames and network metric analysis with the aim to assess how user engagement and message amplification can contribute to influence within the area of social networks and the ways how bot actors can possibly alter the overall dynamics of dissemination. This research is be conducted within the broader framework of communication of the digital political world and computational social science. Recent research in this field could reveal the ways how social media platforms tend to reshape political discourse via mass engagement, algorithmic authorization mechanisms and the disseminationof the information based on the network. In such digitally mediated environments, the political communication is increasingly appearing from certain types of interactions between human users, platform algorithms and automated agents. This study aims to contribute to a truly deeper understanding of the certain ways how the structural characteristics of communication networks can possibly influence the flow and visibility of information within the political context in modern digital public spaces via applying the computational methods to large datasets within the social media environment. Research Contributions This study makes highly valuable contributions to the fields of digital political communication and computational sociology. More specifically, it could conduct a comparative analysis within the computational environment of political communication across two majorly known election cycles, which in result revealed measurable structural changes in the participation of the variability, influence intensity alongside with the network centralization. Second, it could provide a comprehensive analytical framework that enabled to systematically link user activity, message amplification and influence dynamics, which makes a multidimensional assessment of communication processes within the frames of the digital environment possible. Third, it could quantify the amplification effect of accounts considered highly active and potentially automated, which could potentially demonstrate how a presumably tiny number of actors can disproportionately influence the process of wide the information dissemination. Finally, there is a high contribution is to sociology within the computational measurements via demonstrating the potential in order to reveal the underlying structural characteristics of systems regarding the political communication through the ways of integration of large-scale digital tracking data with the feature of simulation modeling. It should be mentioned that this study could surpass previous research in terms of quantitatively monitoring structural changes in the realm of the digital political communication between the period of the indicated election cycles. This study, in particular, indicates that a relatively tiny number of highly active and considered as potentially automated computations can measurably make the information reconstructed in terms of diffusion patterns and network centralization. Using the commonly known ways like integrating behavioral indicators, amplification dynamics and the architectures within network system into a unified computational framework, the current research uses a systematic approach in order to analyze the complex effects in communication environments based on the Twitter platform. Research Questions Based on the theoretical framework and research objectives mentioned above, this study mainly tends to address the following research questions: RQ1: How is political communication distributed across users in election-related discussions on social media, and how do participation patterns shape the production of political content? RQ2: How do amplification mechanisms in digital networks—including retweet dynamics, user network positions, and automated actors—affect the dissemination and visibility of political messages? Hypotheses To empirically test these research questions, the study examines the following hypotheses. Hypothesis 1 Political activity on social media is highly not evenly distributed, with only a small number of users who produce the majority of the political content. Hypothesis 2 Message amplification through retweets can lead to a highly biased dissemination pattern, where a very small percentage of messages can achieve significantly wider reach comparing to the majority of content. Third hypothesis: Accounts that behave automatically exhibit significantly higher levels of activity comparing to the ordinary human users. Fourth hypothesis: Users who take a central position in the social network are inclined to exert greater influence on the dissemination of political information. Fifth hypothesis: The presence of automated bots in the social network can generally increase the potential spread and amplification of political messages. 2. Literature Review and Theoretical Framework 2.1. Digital Political Communication in Platform Societies The transformation within the political communication system in the digital age can be closely linked to the sudden emergence of a communication infrastructure, which is based on digital platforms. Social media platforms like Twitter, Facebook and YouTube are inclined to tuning to into the main centres for the political exchange, which in result let politicians, journalists, activists, and citizens share their information and actively take place in public discourse. Researchers are inclined toward increasingly viewing this transformation via the lens of the platform society, where the production, distribution, and dissemination of information within modern communication systems is actively shaped by digital infrastructure (van Dijck et al., 2018 ; Poell et al., 2019; Gorwa et al., 2020; Kitchin, 2017 ). Taking into account these details, political communication is increasingly considered as a subject to algorithmic systems in which the interaction process and visibility are prioritized and the spread of information online is considered as common. Platform algorithms can currently influence which political messages are visible, how information has been circulating online and which narratives dominate public discourse (Gillespie, 2018 ; Napoli, 2019 ; Bucher, 2018 ; Beer, 2017 ). Therefore, social media platforms are not merely considered as communication tools as they have turned to be structural intermediaries which are able to shape the dynamics of information flow within the politics. The sudden rise of platform communication could fundamentally alter the democratic inclusion and political campaigning. Politicians currently are more likely to rely heavily on digital platforms in order for them to deliver their messages, mobilize supporters and engage directly with the voters. Communication strategies driven on the data, targeted political advertising and campaign tactics that are algorithm-driven are currently playing an increasingly major role in contemporary electoral politics (Kreiss, 2016 ; Karpf, 2024; Bimber, 2014 ; Vaccari, 2013). This shift has been further envisioned via the theory of networked action, which emphasizes the ways how digital network communications enable decentralized forms of political mobilization that is operating outside traditional organizational structures (Bennett & Segerberg, 2012 ; Loader & Mercea, 2011 ; Theocharis et al., 2015 ). In contrast to the traditional model of political communication, where mass media served as the primary gateways to political information (Habermas, 2006 ), the digital environment has enabled individuals to simultaneously create, distribute and then to disseminate political content. The well-known algorithmic frameworks of the platforms for social media, however, is presenting reasonable obstacles to democratic discourse. Via providing more weight to information that in result can fit with the beliefs of users, the recommendation systems fully personalized for users can potentially make ideological divides worse and in result create echo chambers and highly selective patterns in the ways how the chunks of the information is presented (Pariser, 2011 ; Sunstein, 2017 ; Cinelli et al., 2020). Empirical research in this case demonstrates that these dynamics can make political polarization even worse and change drastically typical patterns of political engagement in the digital realm (Iyengar & Hahn, 2009 ; Stroud, 2011 ). Simultaneously, the extensive digital ramifications coming from online interactions present the straight and clear opportunities to be able to analyze communication patterns, ideological networks and the ways of information is disseminated. As a result, social science methods based on the computations are becoming more and more important for studying how digital political communication works (Jungherr et al., 2020 ; Barberá et al., 2015 ). 2.2. Misinformation and Information Disorder The wide spread of misinformation is considered as a central issue in digital communication of the modern world. Information disorder can be commonly divided into 3 categories: misinformation (false information shared without the intent to deceive), disinformation (deliberately misleading content) and malinformation (pretty accurate information used with the aim to cause harm) (Wardle & Derakhshan, 2017 ; Lazer et al., 2018 ; Allcott & Gentzkow, 2017 ). Empirical studies showed that misinformation spreads fast across social media, often spreading even faster than expected and reaching larger audiences than the information identified as the accurate one (Fusoggi et al., 2018). Its circulation is seen to be frequently concentrated within the frames of the ideologically aligned communities that constantly tend to reinforce shared narratives (Grinberg et al., 2019 ; Guess et al., 2018 ). The spread of misinformation is also seen and influenced by known psychological factors like cognitive biases, emotional responses, identity-based reasoning and partisan motivations (Lewandowsky et al., 2017; Pennycook & Rand, 2019 ; van der Linden et al., 2020). Simple interventions, the one is especially known as the prompting users to consider the accuracy before sharing, have been shown to significantly reduce the amount or as a best outcome avoid misinformation sharing (Pennycook et al., 2020). At the structural level, algorithms of the known platform and media systems based on the attention sharing tend to amplify emotionally engaging or the content considerd more sensational, inadvertently increasing the possible visibility of misleading narratives (Benkler et al., 2018 ; Napoli, 2014 ). As a result, these dynamics may collectively contribute to a certain environment where the misinformation gains the high traction and can as an outcome influence public opinion. 2.3. Computational Propaganda and Automated Actors Computational propaganda is considered to be bonded with the wide utilization of automated systems, algorithms and coordinated networks with the final aim to manipulating the public opinion online (Woolley & Howard, 2017 ; Bradshaw & Howard, 2018). A core element of these advanced strategies is involving social bot, which are defined as the automated accounts capable of generating content and interacting with users in order to amplify messages, spread disinformation and in result to increase visibility (Ferrara et al., 2016 ; Howard & Kollanyi, 2016). A plenty of empirical researches have demonstrated that automated accounts are set to actively participating in political discussions and contribute to the pretty active dissemination of unreliable information, particularly taking into account the activity on the platforms like Twitter (X) (Shao et al., 2018 ; Subrahmanian et al., 2016 ). These accounts have the potential to distort information flows via the ways like rapidly reposting content, simulating engagement and coordinating messaging across networks (Stella et al., 2018 ). As a result, detecting automated behavior has become a key focus in computational social science, with the methods of the machine learning developed to identify bots based on behavioral, linguistic and features which are based on the network (Cresci et al., 2017 ; Zhou & Zafarani, 2020). 2.4. Artificial Intelligence and Synthetic Political Communication Modern advancements in the realm of artificial intelligence have led to the introduction of the synthetic media as a new dimension of political communication in the digital world. Generative AI systems currrently have the abilities to producing text, images, audio and video which tend to closely resemble content created by human. A particularly concerning issue nowadays is seen in the development is deepfake technology, which allows the creation of highly realistic and at the same time artificially generated material with the sceptrum of audiovisual info. Content of that ilk shows high risks to democratic discourse via the ways like facilitating the spread of convincing but at the same time the misleading messages in political realm (Vakari & Chatwick, 2020). Studies show that the content generated via AI usage can achieve levels of engagement which can be compared to or exceeding human-generated content, which in result can increas its potential influence. As generative AI becomes more common, the concerns regarding its currrent role in shaping political narratives and public opinion is still seen to actively grow. At the same time, the common detection methods are tend to be developed using machine learning techniques that have the set to analyze visual inconsistencies, behavioral signals and dissemination patterns with the aim to identifing manipulated synthetic media. 2.5. Algorithmic Governance and Democratic Implications The increasing role of systems with the certain algorithms in digital platforms has led to important questions about the possible accountability in terms of governance and democratic. Platform algorithms influence the ways which political content users see, how information can circulate and which narratives currently seen to dominating public discourse (Helberger et al., 2018 ; Kitchin, 2017 ). These systems can often be described as opaque “black boxes,” as it is inclined to limit transparency and make it difficult for users, regulators altogether with the famous researchers with the aim to understanding how information is filtered and distributed (Pasquale, 2015; O'Neil, 2016). This issue can be observed to be closely connected to broader debates on digital political economy, data power and capitalism with the focus on surveillance (Couldry & Mejias, 2019 ; Fuchs, 2017 ). Taking into account the current growing integration of AI into certain communication infrastructures, concerns around transparency, accountability altogether with governance have increasingly seen as a core in contemporary research (Floridi et al., 2018 ; Mittelstadt et al., 2016 ). 2.6. Algorithmic Amplification and Synthetic Influence Framework Based on ideas from platform research, disinformation research and computational propaganda research, the study provides an "algorithmic amplification and artificial influence framework" in order to conceptualize the ways how political messages can gain recognition and influence within systems of digital communication. The Amplification Framework in Digital Political Communication This framework shows the dissemination of information in political realms as a multi-stage amplification process, which can be driven by three interacting mechanisms: Unequal Participation: Political content production can be seen as highly concentrated, with a tiny number of users producing the majority of the current content. This in result reflects the pattern of the power distribution that characterizes networks in the digital world. Algorithmic Amplification: Platform algorithms are inclined to prioritize content related to the high-engagement process (likes, shares, retweets, etc.) and boost its possible visibility through a feedback loop which is actively seen in already popular messages. Artificial Amplification: Automated or semi-automated entities like the social media bots which amplify these dynamics via the ways that generate massive amounts of content, promote specific messages and inflate engagement metrics. The convergence of these highly clever mechanisms creates a communicative environment in which a tiny and not that noteble number of entities can exert a disproportionate influence. Therefore, online political discourse can possibly arise from the interaction between human users, algorithmic system and automated entities. 2.7. Research Gap and Study Contribution Current research seem to indicate that political communication can potentially be influenced by digital platforms, algorithms, automated actors alongside with the techniques related to the artificial intelligence. However, few studies could have addressed these dynamics in the environments of the artificial media via the ways of combining specially selected large-scale election data with the sophisticated analysis of the computational network and behavioral modeling. This study aims to fill this gap via the broad analysis of the political communication and information dissemination in discussions regarding the elections on Twitter (X) using computational social science methods. Via integrating statistical analysis, network modeling, and simulation, author provides empirical insights into the ways of how platform is structured and automated actors can possiblt affect the spread of political discourse. 3.1. Research Design This study seems to adopt a framework of the computational social science with the aim to analyze the structural dynamics within the frames of political communication and artificial amplification on the social media. The research design currentlycombines data analysis with the large-scale dataset, behavioral modeling, network simulations and statistical methods aiming at examining user activity, message amplification and also information diffusion. Via the utilization of large-scale digital trace data taken with the computational modeling allows for investigation on the systematic basic of patterns related to thecommunication that are difficult to capture thorugh using traditional qualitative or small-scale experimental methods. The currently given empirical analysis focuses on Twitter (X), a widely used and well-known platform in studies related to the political communication, disinformation and automated behavior. Its publicly available data like tweets, retweets, replies and engagement metrics making the analysis prcoess in the large scales of communication networks and discourse possible. The study consists of a plenty of computational experiments, which were conducted in MATLAB, each specifically targeting a specific structural aspect of political communication, including: The distribution of user engagement The retweet-driven amplification dynamics The behavioral differences between typical and highly active users The detection of bot-like activity patterns The relationships between engagement metrics The network propagation under the automated amplification Together, these components can form a sophisitcated framework with multiple layers for understanding the methods of how political the information spreading and how automated actors can possibly shape the dynamics of the communication in the environments of the digital world. 3.2. Dataset Construction and Data Sources The given empirical analysis is based on a truly massive dataset of tweets related to the election process posted on Twitter (X) during the 2016 and 2024 US presidential elections. The dataset takes into account data of both tweet-level and user-levels, which allows the analysis of engagement patterns, message dissemination dynamics and the characteristics related to the user behaviour in discussions on political topics. Each of the given observation in the dataset includes multiple variables, which describe communication behavior, such as: Tweet content Timestamp User ID (anonymous) Number of retweets Engagement metrics Hashtag usage patterns The integrated and collected dataset includes approximately 750,000 tweets created by more than the 80,000 users. Specifically, the final dataset is consistsing of: 355,000 tweets from the 2016 election period 402,000 tweets from the 2024 election period These datasets make possible a comparative study of political communication dynamics during the two election periods. The datasets which are analyzed in this study were obtained from a widely-known and publicly available Twitter dataset on the so-called Cagle Research Data Platform. These datasets currently containing tweets with the main topic of election which is collected via the Twitter API and shared for research purposes. They at the same time include tweet-level metadata alike as timestamps, engagement metrics, anonymous user IDs and retweet counts, making a comprehensive analysis of dynamics within the political communication realm in the digital political environment feasible 3.3. Data Preprocessing Prior to the current analysis, the chosen dataset went though several preprocessing procedures in order to guarantee the consistency and reliability in terms of analysis. The preprocessing pipeline included: The removal of duplicate tweets The filtering of incomplete or deleted records The normalization of timestamp formats The verification of temporal consistency across observations The extraction of engagement indicators such as retweet counts User-level activity metrics were subsequently constructed with the aim to measure the participation intensity across the communication network. Previously analysed researches have demonstrated that the participation within the environments of the social media typically follows heavy distributions, in which a relatively not noticeable number of users can lead to the generation of a disproportionately large share of content. Preliminary inspection of the previously chosen dataset confirmed that patterns of the similar participation are present in the empirical data are analysed in this study. 3.4. Behavioral Modeling of Communication Activity The behavior of the users was analyzed via utilization of metrics based on the activity, which is derived from the dataset, taking into account the number of tweets per user, retweet rate, engagement intensity and the index of a composite activity. The idea of the index is to combine posting frequency and retweet activity with the goal to capture overall communication intensity in the discourse of the politics. Chosen users above the 95th percentile of activity were inclnied to be classified as highly active. This classification can not directly identify bots but at the same time it follows a common approach in behavioral studies within the computational social science where bot detection tools are not that widely available. Such accounts with the high activity may include automated agents, coordinated accounts or even human users with presumably high engagement. 3.5. Synthetic Network Construction Taking into acount the goals of examining self-amplification effects, a network of the artificial communication was constructed with the full implementation of a scale-free (non-scale) model, which is based on preferential attachment. This approach can reflect social media structures seen and identified in the real-world, in which highly connected users tend to attract more interactions. The network was in result initialized with 3 fully connected nodes, with additional nodes added step by step based on the preferential attachment (i.e., connection probability proportional to node connectivity). The initial network of the experiment included 600 nodes taken to analyze structural properties in a controlled environment. Larger simulations were conducted with 5,000 nodes created to test robustness, particularly with the aim of analyzing automated amplification dynamics (see Section 4.7 ). Even though it is smaller than real-world networks, these simulations make key structural features viisble and give approximate meanings and provide a controlled framework with studying the effects of self-amplification in the systems of the digital communication. 3.6. Modelling Automated Amplification To investigate the possible impact of actors with full automatization, bots were introduced into an artificial network. Th conducted by author simulation compares two experimental scenarios: The base network (consisting only of human users) The enhanced network (with the addition of automated bot accounts) In the enhanced scenario, automated actions were actually distributed within approximately 400 nodes, which are based on the previously mentioned rating showing the high activity. The bots were designed as nodes showing high-activity that actively created and retweeted content and also mirroring the actual activity patterns that are observed in the dataset. 3.7. Bot Amplification Mechanism Author has simulated an auto-amplification strategy with the way of allowing bot nodes in order to create additional links beyond the underlying network structure. For each bot node: Between 5 and 10 additional links were created. 70% of the new links targeted higher-order nodes (network hubs). 30% of the links targeted randomly selected users. This mechanism could successfully mimic the strategies of auto-amplification, which are common in online political networks, in which automated accounts can increase the probable visibility of messages with the way of frequently interacting with influential users and trending topics. 3.8. Network Metrics In order to assess the structural differences between organic and amplified communication networks, author could calculat several network metrics: Mean score: It shows the average number of connections per node. Extreme score: It indicates the influence of necessary nodes with a plenty of connections within the network. Aggregation coefficient: This indicates the tendency of nodes with the goal to form closely interconnected local clusters. Furthermore, author also investigated the following distributions: Score distribution Score centrality distribution Via utilization of these metrics, author was able to conduct a structural comparison between organic and communication networks with the self-amplification. 3.9. Correlation Analysis With the aim to analyze the relationships between communication measures, a correlation analysis was conducted on multiple interaction variables. The variables analyzed included: Tweet frequency Retweet count Interaction intensity An impact index calculated from the interaction measures. Pearson's correlation coefficient was implemented in order to assess the statistical relationships between the chosen variables. 3.10. Experimental Implementation All given computational experiments were conducted through the MATLAB program, which provides a big number of tools for numerical calculations, statistical modelling and graphics. The experimental framework utilized the following tools: Statistical analysis functions Numerical modeling tools for simulating behavior Graphalographic tools for representing multidimensional data The MATLAB scripts could generate the graphs illustrating the structural patterns of communication within the political envirnoment, including activity distribution, retweet amplification dynamics and the processes of the simulated network propagation. All provided graphs were exported at 600 dots per inch (DPI) to meet the standards of academic publication. 3.11. Statistical Significance The statistical significance of the observed differences was evaluated via using the method of the standard inference. All statistical tests were made at a 95% confidence level and following results with a p-value less than 0.05 could be considered statistically significant. Additional tests were conducted with the aim to verify the robustness of the results, making possible the fact that the observed communication patterns were not influenced by individual observations or any other factors. The consistency of the relationships with certain structures in the two election datasets could support the stability of the experimental results. 3.12. Methodological Limitations First thing to be noted, the analysis relies on Twitter (X) data, which can represent only a bordered segment of the broader landscape into the political communication. Political discourse may also occur through other channels, taking into account traditional media, private messaging platforms and interactions made offline that are not captured in public datasets. Second, the identifying automated or coordinated accounts which is based solely on behavioral metrics is considerd as inherently difficult. High activity levels may indicate automation, but at the same time they can also possibly reflect highly engaged human users. Additionally, social media platforms frequently tend to changing their algorithms and policies of the moderation, which can in result influence communication patterns and information dissemination over specified period of time. Given these constraints, the findings should be interpreted with caution. The applied metrics are useful with the aims of detecting extreme activity patterns but do not provide the full definitive identification of automated actors. Therefore, the results are not consideredv as fully generalizable through all platforms or communication environments. 3.13. Reproducibility and Transparency To be able to ensure transparency and reproducibility, all computational experiments were conducted with the utilization of the standardized MATLAB software that also simulates the analytical workflow, which is described in this study. This software includes the following steps: raw data processing, statistical analysis, behavioral modeling and network simulation. This computational workflow could provide a transparent methodological framework that make possible for future researchers interested in the topic to replicate the analytical procedures under equivalent conditions for the experiment. 3.14. Originality of Figures and Tables All figures, tables, graphs, and simulation outputs, which are presented in this study were generated by the author using the computational methods and datasets, that are described in the Methodology section. No figures or tables were reproduced from external publications or electronic sources. All graphs were specifically designed for this study by implementation of the computational analysis based on MATLAB software. 4. Results 4.1. Overview of Experimental Design A number of computational analyses using MATLAB tool was conducted for examining the characteristics like user engagement, message amplification, highly active accounts and network structures in political communication within the digital world. The study is focused on the presidential elections between 2016 and 2024 US, providing a perspective on comparison on the evolution of political discourse online. The datasets include approximately 355,000 tweets from 82,417 users (2016) and 402,000 tweets from 96,210 users (2024). The impact scores were calculated through the usage of a composite index based on retweet rates, follower interactions and message propagation within the given network. All current figures and tables are derived from the described analysis on the computational background. The results show that the political communication is considered as highly concentrated: a unnoticeable number of highly active accounts tend to generate and amplify a large share of the overall content. Simulation findings further indicate that the automated or highly active actors are more likely to significantly increase the network propagation, which is also highlighting the algorithmic amplification’s importance in shaping the digital political discourse. 4.2. Distribution of User Activity The first experiment has examines the distribution of tweets among users, which are participating in election related to discussions. Engagement patterns on social media platforms are generally seen and considered as highly not even, with only given small number of users accounting for the majority of communication activities. Table 1 Dataset Summary Statistics Metric 2016 Election 2024 Election Total Tweets 355,000 402,000 Total Users 82,417 96,210 Avg Tweets per User 4.31 4.18 Avg Retweets per Tweet 6.8 8.9 Median Retweets 2 3 High-Activity Users 4,128 6,214 % High-Activity Users 5.0% 6.5% The 2016 election data is seen to be contained 355,000 tweets from 82,417 users, which is averaging 4.31 tweets per user. However, user engagement levels are likely to be varied considerably. Approximately 68.4% of users posted fewer than 3 tweets comparing to the percentage of 24.4% posted between 23 and 24.4% posted 20 or more tweets. This suggests that a relatively tiny number participants showing high activity had a disproportionately large impact on the overall political discourse’ volume. A similar trend could be observed in the 2024 election data. It contained 402,000 tweets from 96,210 users, averaging 4.18 tweets per user. Approximately 71.6% of users seemed to post fewer than three tweets, at the same time 22.1% posted between 23 and 24.4% posted 20 or more tweets. These results revealed a striking long-tailed engagement structure, where a small number of users show a major portion of the political discourse. This visual representation shows a significantly high imbalance in the engagement’s structure, with a small percentage of users who are constituting the majority of communication within the political frame. Overall, these results tend to provide strong empirical evidence, which are supporting the first hypothesis, which posits that activity related to the political communication on social media follows a highly unequal distribution of engagement. 4.3. Retweet Amplification Dynamics The second experiment shows the behavior of retweeting from the perspective of a mechanism for amplifying messages in the networks of the political communication. This figure actively shows the relationship between the number of retweets, user engagement and message reaching across three dimensions. In the 2016 election data, the average number of retweets was 6.8 (standard deviation 14.2), with a median of 2. The distribution of retweets was highly skewed. Approximately 74.3% of tweets received the number fewer than 5 retweets, while 21.1% received between 5 and 50 retweets. Only 4.6% of tweets received more than 50 retweets, which is indicating that a small number of messages achieved relatively high levels of reach. During the 2024 election period, the dynamics of reach tended to increase. The average number of retweets was 8.9 (standard deviation 18.7), with a median of 3. Approximately 69.8% of tweets received fewer than 5 retweets, 24.5% received between 5 and 50 retweets, and 5.7% received more than 50 retweets. These results indicate a high improvement in message dissemination and communication between the two election periods. 4.4. Detection of Bot-Like Behavioral Patterns In the third experiment, author had examined the behavioral characteristics, which are associated with accounts that might probably operate automatically or in coordination with other given accounts. With the aim to identify these patterns, the author calculated a composite activity index, which was combining the frequency of tweets and retweet engagement metrics. This visual representation indciates that regular users and highly active accounts are seemed to categorized within a behavioral space in the frames of three dimensions, that is defined by tweet frequency and retweet engagement patterns. Accounts being at above the top 95th percentile of the activity distribution were categorized as exhibiting the patterns of the bot-like behavior. In the election data of 2016, approximately 4,128 accounts (5% of users) showed unusually high levels of activity. These accounts posted an average of 36.4 tweets per user, which seem to be significantly higher than the average of approximately 3.1 tweets per user. Furthermore, highly active accounts achieved an average of 21.7 retweets per tweet, comparing it to approximately 5.9 retweets per user. Applying the same categorization procedure to the 2024 data identified approximately 6,214 accounts (6.5% of users) with similar behavioral characteristics. These results are likely to indicate a gradual increase in the number of accounts showing high activity, or accounts that are likely to be automated, between the two election cycles. 4.5. Behavioral Threshold Structure In order to further investigate differences in behaviour within the system of communication, we created a activity space of three-dimensions that combines tweet frequency, retweet participation and composite activity score. This graphical picture shows the distribution of the activity in behavioural measures and the thresholds used aiming at distinguishing between typical engagement patterns and markedly elevated activity related to behaviour. To differentiate the typical user behaviour from the one considered as abnormally high, behavioural thresholds which are corresponding to the 95th percentile of the activity distribution were used. 4.6 Correlation Between Activity and Message Amplification Author investigated the relationship between user activity and message amplification with the utilization of correlation analysis of key communication metrics. Table 2 Correlation Results Variable Pair Correlation (r) Tweets vs Retweets 0.47 Retweets vs Influence 0.71 The analysis in result revealed a relatively moderate to strong positive relationship between communication activities, message amplification and influence. This graph illustrates the relationships between user activity, engagement levels, and impact measures. In order to further analyze these relationships, author estimated a multiple regression model. Influence = β0 + β1(Activity) + β2(Retweets) This model can explain a large part of the dispersion in the user’s influence (R² = 0.58, p < .001), and shows that activity level and retweet amplification are considered strong indicators of influence within the communicational network. 4.7. Bot-Driven Amplification Experiment To assess the probable impact of automated actors on the dissemination of the information under certain large-scaled conditions, a simulation experiment was conducted with the using an expanded artificial communications network with 5,000 users. This large-scale network is seen as an extension of the basic structural model, which is described in Section 3.5 and enables the analysis of amplification dynamics in communication environments considered more complex. This graph illustrates the accelerated spread of messages due to the usage of automated accounts. Two scenarios were analyzed: A baseline network which is consisting solely of human users. A network comprising 400 automated accounts (bots). The results showed that human users retweeted an average of 12.7 times per message, while in comparison at the same time automated accounts retweeted an average of 94.2 times per message. An independent samples t-test confirmed that this difference was statistically significant. t(5398) = 18.64, p < 0.001. The use of automated accounts increased the spread of information across the entire network by approximately 38%. 4.8. Network Structural Effects of Bot Amplification Table 3 summarizes the structural characteristics of the simulated communication network. Table 3 Network metrics Metric No Bots Bots Average Degree 5.98 7.40 Max Degree 92 97 Clustering Coefficient 0.050 0.061 Network Density 0.0024 0.0031 Average Path Length 4.62 4.21 The results indicate that auto-amplification actually improves network connectivity overall with a slight enhancement of local aggregation structures. Figure 8 illustrates the configuration of the network after the automated computing system is implemented. The red nodes represent the bot programs, while the blue ones represent the human users. The independent samples t-test, when comparing the distributions of node scores could yield a p-value of 0.00257, which is indicating a statistically significant difference between the two network configurations. Figure 9 shows that the communication network arrangement follows a thick-tailed distribution. The presence of bots increases the density of nodes with a large number of connections. The effect size (Cohen's coefficient d = 0.20) indicates that the self-amplification has a significant effect, even in the fact if it is structurally small. Figure 10 illustrates the ranking of node centrality in scenarios of the both network, demonstrating that the automated amplification can enhance the importance of already influential nodes. Figures 7 – 9 depict the structure of the communication network, score distribution and centrality ranking in scenarios related to both networks. Overall, these results indicate that automated actors improved the ability to disseminate political messages via enhancing the coherence of the network and strengthening the hub structure within the communication system. 4.9. Comparative Analysis of Election Cycles Comparative analysis revealed that the structural patterns showing several consistencies across two election cycles. First, political engagement remains the highly concentrated realm, with a small number of active users who are producing the majority of content related to election. Second, the intensity of message dissemination could increase between the two election cycles, which is indicating a strengthening of the dissemination of information dynamics within the network. Third, the proportion of active accounts or accounts likely to be automated has gradually increased over the time. These findings provide the empirical evidence with the goal to answer research questions four and five, demonstrating that the automated actors can change the structural characteristics regarding the political social networks and enhance their capabilities of information dissemination. 4.10. Influence Density Landscape of Digital Political Communication To incorporate the structural dynamic properties revealed in previous experiments, a three-dimensional model of impact intensity was created. This graph can clearly illustrate the differences in engagement, information dissemination intensity and automated account activity between the two election cycles of 2016 and 2024 years. Furthermore, this graph demonstrates the ways how user interaction with message dissemination contributes to the possible creation of concentrated spheres of influence within communication networks. In Fig. 12 peak intensity refers to the area within the system of the political communications where the influence is most concentrated. 4.11. Galaxy Visualization of Political Communication Networks To illustrate the structural complexity of the environment for the political communication, a three-dimensional representation of the assemblies was created. In Fig. 13 the three-dimensional coordinates represent the level of the standard activity, and degrees of participation and influence are taken from the interaction indicators. The resulting cluster structure indicates the potential emergence of distinct communities and centers of influence within the network of the political communication. 5. Discussion This study examined of the digital communication in the political realm via using large-scale data driven form the social media and the analysis of the computational network of the presidential elections related to 2016 and 2024 years. The results show that online discourse in politics is shaped by unequal participation, amplification mechanisms, network centralization and automated activity, which is supporting prior research on the transformations on the structural level in networked communication (Tucker et al., 2018 ; Schillemans, 2014 ). Even though the early perspectives emphasized the potential of democratizing in digital platforms, the findings indicate the sharply increasing concentration. A tiny group of highly active users takes the functions of s central nodes, with algorithmic amplification reinforcing their visibility and influence within the network. 5.1. Concentration of Political Communication Political communication seems to be highly concentrated among only a small number of users. In 2016, the top 1% produced 18.7% of tweets, increasing to 21.4% in 2024, which is truly indicating growing centralization. These patterns align with power-law distributions in digital networks (Barabasi, 2016; Newman, 2018), where influence is concentrated among the so-called central actors. This suggests that theonline political discourse is shaped by the participation inequalities, which in result is leading to emerging power hierarchies rather than fully decentralized communication structures. 6. Conclusion This study has made a full analysis of political communication through large-scale data and network modeling, focusing on how engagement, amplification and automation can potentially influence the diffusion of information. According to the findings the communication is highly concentrated, amplification is not even and only a small number of messages and users dominate the main visibility. Bot-like accounts, even though they limited in number, contribute disproportionately to the process of content production and spread, while network structures seems to remain strongly centralized. Comparative results are pointing at the intensity on the increasing amplification and the centralization over time. Simulations further demonstrate that even a small numbers of automated actors can significantly enhance the diffusion of the messages. Overall, political communication within the digital world is driven by interconnected amplification processes involving users, algorithms and automated systems, which is leading to concentrated influence. These results support the framework of algorithmic amplification and artificial influence, which is at the same time highlighting important implications for platform governance, transparency and democratic resilience. Future research should extend to cross-platform dynamics, improve the accurate detection of automated actors and further examine the role of AI in shaping the political communication. Glossary Algorithmic Amplification - The known process by which the platform algorithms enhancing the visibility of content through the systems of recommendation and mechanisms of the ranking. Algorithmic Amplification – The process by which algorithms of platform increasing the visibility of content through systems of recommendation and ranking mechanisms. Amplification – The process of increasing the visibility and reach of the content within the digital world through reposting, retweeting, algorithmic promotion or collaborative activities. Automated Account - A social media account which is managed partially or entirely by scripts based on automation, bots, or algorithmic systems, rather than through the ones with direct human interaction. Automated Account – Accounts of social media that are managed partially or entirely by automated scripts, bots or algorithmic systems, rather than through the ones with direct human interaction. Bot Activity – Observable behavioral patterns that are produced by automated accounts. These include posting frequency, retweeting behavior and collaborative amplification. Computational Communication Model – A systematic framework with the aim of simulating and evaluating communication dynamics by integrating the empirical data in social media with computational analysis. Digital Political Communication – The exchange of information, opinions and political discourse across various platforms in online world and networks of social media. Echo Chamber – A communication environment where individuals are primarily exposed to information that reinforces their existing beliefs. Engagement Metrics – Quantitative metrics used with the aim to measure the interaction of user with digital content (such as likes, retweets, replies, and shares). Influence Score – A composite metric used with aim to assess the impact of an account within a social network. It is calculated based on engagement frequency, sharing patterns and the reach itself. Information Diffusion – The process of messages, information and opinions that are spreading across a network through the process of user interactions, such as sharing, reposting and reacting. Misinformation – The information considered as false or misleading that spreads within a digital communication network, not depending on the intent. Network Centrality - A structural property of nodes within a network, that is indicating their relative importance or influence within the structure of communication. Network Centrality – The structural characteristics of nodes within a network, that are indicating the relative importance or influence of nodes in the structure of communication. Declarations Author Contributions The author was fully responsible for conceptualization, methodology development, data collection, formal analysis, visualization and preparation of the original manuscript as he prepared it by himself. Funding This research received no external funding. Ethics Statement This study relies exclusively on publicly available data of social media and does not include any information considered as personally identifiable or personally identifiable, except for publicly available user IDs. This study adheres to the ethical guidelines that are adopted in computational social science and the guidelines for the responsible use of data ofdigital tracking. Reproducibility Statement The computational experiments were performed via the utilization of the analytical software that is based on MATLAB. The systematic workflow and analysis procedures were described in sufficient detail to allow their reproduction via using equivalent datasets. Data Availability Statement The datasets analyzed in this study are conisdered as publicly available through the Kaggle Research Data Platform, which is a publicly accessible Twitter dataset highly related to elections. This data was originally collected using the Twitter API and is widely used in academic research on communication in digital political. Conflict of Interest The author declares no conflict of interest. Acknowledgments The authors extend their sincere thanks to their colleagues and reviewers for their valuable feedback, which contributed to improving the quality and clarity of this study. They also express their deep gratitude to everyone who contributed data, materials, and other resources necessary for the completion of this research. References Acemoglu D, Restrepo P (2020) Artificial intelligence, automation, and work. J Economic Perspect 34(1):3–30. https://doi.org/10.1257/jep.34.1.3 Allcott H, Gentzkow M (2017) Social media and fake news in the 2016 election. J Economic Perspect 31(2):211–236. https://doi.org/10.1257/jep.31.2.211 Allcott H, Gentzkow M, Yu C (2019) Trends in the diffusion of misinformation on social media. Res Politics. https://doi.org/10.1177/2053168019848554 Altay S, Nielsen RK, Fletcher R (2022) Quantifying the relationship between misinformation exposure and media use. New Media Soc. https://doi.org/10.1177/14614448221091216 Aral S, Eckles D (2019) Protecting elections from social media manipulation. Science 365(6456):858–861. https://doi.org/10.1126/science.aaw8243 Badawy A, Addawood A, Lerman K, Ferrara E (2019) Characterizing the 2016 Russian IRA influence campaign. Social Netw Anal Min 9., Article 1. https://doi.org/10.1007/s13278-019-0578-6 Badawy A, Lerman K, Ferrara E (2018) Who falls for online political manipulation? The Russian Twitter campaign during the 2016 U.S. election. In Proceedings of the World Wide Web Conference (pp. 162–168). https://doi.org/10.1145/3308560.3316494 Barberá P (2015) Birds of the same feather tweet together: Bayesian ideal point estimation using Twitter data. Political Anal 23(1):76–91. https://doi.org/10.1093/pan/mpu011 Barberá P, Jost JT, Nagler J, Tucker JA, Bonneau R (2015) Tweeting from left to right: Is online political communication more than an echo chamber? Psychol Sci 26(10):1531–1542. https://doi.org/10.1177/0956797615594620 Beer D (2017) The social power of algorithms. Inform Communication Soc 20(1):1–13. https://doi.org/10.1080/1369118X.2016.1216147 Bender EM, Gebru T, McMillan-Major A, Shmitchell S (2021) On the dangers of stochastic parrots: Can language models be too big? In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency (pp. 610–623). https://doi.org/10.1145/3442188.3445922 Benedetto A (2025), September 22 Generative AI, disinformation, and electoral integrity: The EU's response to democratic challenges. ITSS Verona Magazine, 4 (1). https://www.itssverona.it/itss-magazine Benkler Y (2019) Networked propaganda and disinformation. Harvard Kennedy School Working Paper . https://doi.org/10.14195/2183-5462_34_24 Benkler Y, Faris R, Roberts H (2018) Network propaganda: Manipulation, disinformation, and radicalization in American politics. Oxford University Press. https://doi.org/10.1093/oso/9780190923624.001.0001 Bennett WL, Segerberg A (2012) The logic of connective action. Inform Communication Soc 15(5):739–768. https://doi.org/10.1080/1369118X.2012.670661 Bessi A, Ferrara E (2016) Social bots distort the 2016 U.S. election online discussion. First Monday 21(11). https://doi.org/10.5210/fm.v21i11.7090 Bimber B (2014) Digital media in the Obama campaigns of 2008 and 2012: Adaptation to the personalized political communication environment. J Inform Technol Politics 11(2):130–150. https://doi.org/10.1080/19331681.2014.895691 Bommasani R, Hudson DA, Adeli E, Altman RB, Arora S, von Arx S, Bernstein MS, Bohg J, Bosselut A, Brunskill E, Brynjolfsson E, Buch S, Card D, Castellon R, Chatterji N, Chen A, Creel K, Davis JQ, Demszky D, Ladhak F (2021) On the opportunities and risks of foundation models. arXiv. https://doi.org/10.48550/arXiv.2108.07258 Bovet A, Makse HA (2019) Influence of fake news in Twitter during the 2016 election. Nat Commun 10 Article 7. https://doi.org/10.1038/s41467-018-07761-2 Bradshaw S, Howard PN (2019) The global disinformation order: 2019 global inventory of organised social media manipulation (Working Paper 2019.2). Oxford Internet Institute, University of Oxford. https://digitalcommons.unl.edu/scholcom/207 Brown TB, 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 DM, Wu J, Amodei D (2020) Language models are few-shot learners. arXiv. https://doi.org/10.48550/arXiv.2005.14165 Bucher T (2018) If… then: Algorithmic power and politics. Oxford University Press. https://doi.org/10.1093/oso/9780190493028.001.0001 Carlson M (2015) The robotic reporter: Automated journalism. Digit Journalism 3(3):416–431. https://doi.org/10.1080/21670811.2014.976412 Cinelli M, De Francisci Morales G, Galeazzi A, Quattrociocchi W, Starnini M (2021) The echo chamber effect on social media. Proceedings of the National Academy of Sciences, 118 (9), e2023301118. https://doi.org/10.1073/pnas.2023301118 Cinus F, Minici M, Luceri L, Ferrara E (2024) Exposing cross-platform coordinated inauthentic activity in the run-up to the 2024 U.S. election. arXiv. https://doi.org/10.48550/arXiv.2410.22716 Conover MD, Ratkiewicz J, Francisco M, Gonçalves B, Flammini A, Menczer F (2011) Political polarization on Twitter. In Proceedings of the International AAAI Conference on Web and Social Media . https://doi.org/10.1609/icwsm.v5i1.14126 Couldry N, Mejias UA (2019) The costs of connection: How data is colonizing human life and appropriating it for capitalism. Stanford University Press. https://doi.org/10.1515/9781503609754 Crawford K, Dobbe R, Fried G, Kaziunas E, Kak A, Mathur V, Richardson R, Schultz J, Schwartz O, West SM, Whittaker M (2018) AI Now 2018 report. AI Now Institute Cresci S, Di Pietro R, Petrocchi M, Spognardi A, Tesconi M (2017) The paradigm shift of social spambots. In Proceedings of the World Wide Web Conference (pp. 963–972). https://doi.org/10.1145/3041021.3055135 Dai Z, Yang Z, Yang Y, Carbonell J, Le QV, Salakhutdinov R (2019) Transformer-XL: Attentive language models beyond a fixed-length context. arXiv. https://doi.org/10.48550/arXiv.1901.02860 Das D, Clark A (2019) Satire vs fake news: You can tell by the way they say it. IEEE Trans Artif Intell. https://doi.org/10.1109/TransAI46475.2019.00012 Davis CA, Varol O, Ferrara E, Flammini A, Menczer F (2016) BotOrNot: A system to evaluate social bots. In Proceedings of the WWW Companion Conference (pp. 273–274). https://doi.org/10.1145/2872518.2889302 Ecker UKH, Lewandowsky S, Cook J, Schmid P, Fazio LK, Brashier N, Kendeou P, Vraga EK, Amazeen MA (2022) The psychological drivers of misinformation belief and its resistance to correction. Nat Reviews Psychol 1(1):13–29. https://doi.org/10.1038/s44159-021-00006-y Enli G (2017) Mediated authenticity in political communication. Inform Communication Soc 20(8):1218–1233. https://doi.org/10.1080/1369118X.2016.1209330 Ferrara E (2017) Disinformation and social bot operations in the run-up to the 2017 French presidential election. Commun ACM 60(7):33–35. https://doi.org/10.1145/3097340 Ferrara E, Varol O, Davis C, Menczer F, Flammini A (2016) The rise of social bots. Commun ACM 59(7):96–104. https://doi.org/10.1145/2818717 Floridi L, Cowls J, Beltrametti M, Chatila R, Chazerand P, Dignum V, Luetge C, Madelin R, Pagallo U, Rossi F, Schafer B, Valcke P, Vayena E (2018) AI4People—An ethical framework for a good AI society: Opportunities, risks, principles, and recommendations. Mind Mach 28(4):689–707. https://doi.org/10.1007/s11023-018-9482-5 Freelon D, Wells C (2020) Disinformation as political communication. Political Communication 37(2):145–156. https://doi.org/10.1080/10584609.2020.1723755 Fuchs C (2017) Social media: A critical introduction, 2nd edn. Sage Gillespie T (2018) Custodians of the internet: Platforms, content moderation, and the hidden decisions that shape social media. Yale University Press. https://doi.org/10.12987/9780300235029 Goodfellow I, Bengio Y, Courville A (2016) Deep learning . MIT Press. ISBN 0262035618 Graefe A (2016) Guide to automated journalism . Tow Center for Digital Journalism. https://doi.org/10.7916/D80G3XDJ Grinberg N, Joseph K, Friedland L, Swire-Thompson B, Lazer D (2019) Fake news on Twitter during the 2016 election. Science 363(6425):374–378. https://doi.org/10.1126/science.aau2706 Guess AM, Nagler J, Tucker JA (2019) Less than you think: Prevalence and predictors of fake news dissemination on Facebook. Sci Adv 5(1):eaau4586. https://doi.org/10.1126/sciadv.aau4586 Guess A, Nyhan B, Reifler J (2018), January 9 Selective exposure to misinformation: Evidence from the consumption of fake news during the 2016 U.S. presidential campaign . European Research Council Habermas J (2006) Political communication in media society: Does democracy still enjoy an epistemic dimension? Communication Theory 16(4):411–426. https://doi.org/10.1111/j.1468-2885.2006.00280.x Helberger N, Pierson J, Poell T (2018) Governing online platforms: From contested to cooperative responsibility. Inform Soc 34(1):1–14. https://doi.org/10.1080/01972243.2017.1391913 Hongladarom S (2023) [Review of the book The age of surveillance capitalism: The fight for a human future at the new frontier of power , by S. Zuboff]. AI & Society, 38 (6), 2359–2361. https://doi.org/10.1007/s00146-020-01100-0\ Howard PN (2020) Lie machines: How to save democracy from troll armies, deceitful robots, junk news operations, and political operatives . Yale University Press. ISBN 978-0300250206 Howard PN, Woolley S, Calo R (2018) Algorithms, bots, and political communication in the US 2016 election: The challenge of automated political communication for election law and administration. J Inform Technol Politics 15(2):81–93. https://doi.org/10.1080/19331681.2018.1448735 Iyengar S, Hahn KS (2009) Red media, blue media: Evidence of ideological selectivity in media use. J Communication 59(1):19–39. https://doi.org/10.1111/j.1460-2466.2008.01402.x Jungherr A (2016) Analyzing political communication with digital trace data. Springer. https://doi.org/10.1007/978-3-319-20319-5 Jungherr A (2023) Digital campaigning: How digital media change the work of parties and campaign organizations and impact elections. Edw Elgar. https://doi.org/10.4337/9781789906769.00035 Jungherr A, Rivero G, Gayo-Avello D (2020) Retooling politics: How digital media are shaping democracy. Cambridge University Press. https://doi.org/10.1017/9781108297820 Just N, Latzer M (2017) Governance by algorithms: Reality construction by algorithmic selection on the internet. Media Cult Soc 39:238–258. https://doi.org/10.1177/0163443716643157 Keller FB, Schoch D, Stier S, Yang J (2020) Political astroturfing on Twitter: How to coordinate a disinformation campaign. Political Communication 37(2):256–280. https://doi.org/10.1080/10584609.2019.1661888 Kitchin R (2017) Thinking critically about and researching algorithms. Inform Communication Soc 20(1):14–29. https://doi.org/10.1080/1369118X.2016.1154087 Kreiss D (2016) Prototype politics: Technology-intensive campaigning and the data of democracy (1st ed.). Oxford University Press. ISBN 978-0199350254 Kreiss D, Lawrence RG, McGregor SC (2018) In their own words: Political practitioner accounts of strategic social media use. Political Communication 35(1):8–31. https://doi.org/10.1080/10584609.2017.1334727 Kshetri N, Voas J (2017) The economics of fake news. IT Prof 19(6):8–12. https://doi.org/10.1109/MITP.2017.4241459 Lazer DMJ, Baum MA, Benkler Y, Berinsky AJ, Greenhill KM, Menczer F, Metzger MJ, Nyhan B, Pennycook G, Rothschild D, Schudson M, Sloman SA, Sunstein CR, Thorson EA, Watts DJ, Zittrain JL (2018) The science of fake news. Science 359(6380):1094–1096. https://doi.org/10.1126/science.aao2998 LeCun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521:436–444. https://doi.org/10.1038/nature14539 Loader BD, Mercea D (2011) Networking democracy? Social media innovations and participatory politics. Inform Communication Soc 14(6):757–769. https://doi.org/10.1080/1369118X.2011.592648 Metaxas PT, Mustafaraj E, Gayo-Avello D (2011) How (not) to predict elections using social media. In Proceedings of the IEEE International Conference on Privacy, Security, Risk and Trust . https://doi.org/10.1109/PASSAT/SocialCom.2011.98 Mittelstadt BD, Allo P, Taddeo M, Wachter S (2016) The ethics of algorithms: Mapping the debate. Big Data Soc 3(2):1–21. https://doi.org/10.1177/2053951716679679 Mutz DC (2006) Hearing the other side: Deliberative versus participatory democracy. Cambridge University Press. https://doi.org/10.1017/CBO9780511617201 Napoli PM (2014) Automated media: An institutional theory perspective on algorithmic media production and consumption. Communication Theory 24(3):340–360. https://doi.org/10.1111/comt.12039 Napoli PM (2019) Social media and the public interest. Columbia University. https://doi.org/10.7312/napo18454 Nieborg DB, Poell T, van Dijck J (2025) Platformization. Edward Elgar. https://doi.org/10.4337/9781035301447.vol3.00051 Nyhan B (2020) Facts and myths about misinformation. J Economic Perspect 34(3):220–236. https://doi.org/10.1257/jep.34.3.220 Nyhan B, Reifler J (2015) The effect of fact-checking on misinformation. Political Communication 32(1):127–150. https://doi.org/10.1080/10584609.2014.969639 O’Neil C (2016) Weapons of math destruction: How big data increases inequality and threatens democracy . Crown Publishers. ISBN 978-0553418811 Oyekunle D, Matthew U, Preston D, Boohene D (2024) Trust beyond technology algorithms: A theoretical exploration of consumer trust and behavior in technological consumption and AI projects. J Comput Commun 12:72–102. https://doi.org/10.4236/jcc.2024.126006 Pariser E (2011) The filter bubble: What the Internet is hiding from you . Penguin Press. ISBN 978-0-670-92038-9 Pennycook G, Rand DG (2019) Fighting misinformation on social media using crowdsourced judgments of news source quality. Proceedings of the National Academy of Sciences, 116 (7), 2521–2526. https://doi.org/10.1073/pnas.1806781116 Porlezza C (2019) Book review: Nicholas Diakopoulos Automating the news: How algorithms are rewriting the media . Journalism 20(10):1417–1419. https://doi.org/10.1177/1464884919876179 Prior M (2007) Post-broadcast democracy: How media choice increases inequality in political involvement and polarizes elections . Cambridge University Press. ISBN 978-0-521-67533-8 Prior M (2007) Post-broadcast democracy: How media choice increases inequality in political involvement. Cambridge University Press Ramanaharan R, Guruge DB, Agbinya J (2025) Deepfake video detection: Insights into model generalisation—A systematic review. Data Inform Manage 9 Article 100099. https://doi.org/10.1016/j.dim.2025.100099 Roozenbeek J, van der Linden S (2019) Fake news game confers psychological resistance against online misinformation. Palgrave Commun 5., Article 65. https://doi.org/10.1057/s41599-019-0279-9 Russell S, Norvig P (2021) Artificial intelligence: A modern approach (4th ed.). Pearson. ISBN 978-0-13-461099-3 Sayers JG (2014) Book review: Communication power by M. Castells. Work Employ Soc 28(1):142–144. https://doi.org/10.1177/0950017013511515 Schillemans T (2014) The hybrid media system: Politics and power. Public Adm 92(4):867–883. https://doi.org/10.1111/padm.12108 Shao C, Ciampaglia GL, Varol O, Yang K-C, Flammini A, Menczer F (2018) The spread of low-credibility content by social bots. Nat Commun 9:4787. https://doi.org/10.1038/s41467-018-06930-7 Smith D, Ehrett C, Warren P (2025) Unsupervised detection of coordinated information operations in the wild. EPJ Data Sci 14(1):26. https://doi.org/10.1140/epjds/s13688-025-00544-y Starbird K (2017) Examining the alternative media ecosystem through the production of alternative narratives of mass shooting events on Twitter. In Proceedings of the International AAAI Conference on Web and Social Media . https://doi.org/10.1609/icwsm.v11i1.14878 Stella M, Ferrara E, De Domenico M (2018) Bots increase exposure to negative and inflammatory political content in online social systems. Proceedings of the National Academy of Sciences, 115 (49), 12435–12440. https://doi.org/10.1073/pnas.1803470115 Stroud NJ (2011) Niche news: The politics of news choice. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199755509.001.0001 Subrahmanian VS, Azaria A, Durst S, Kagan V, Galstyan A, Lerman K (2016) The DARPA Twitter bot challenge. Computer 49(6):38–46. https://doi.org/10.1109/MC.2016.183 Sunstein CR (2017) #Republic: Divided democracy in the age of social media. Princeton University Press. https://doi.org/10.2307/j.ctv8xnhtd Theocharis Y, Lowe W, van Deth JW, García-Albacete GM (2015) Using Twitter to mobilize protest action. Inform Communication Soc 18(2):202–220. https://doi.org/10.1080/1369118X.2014.948035 Tucker JA, Guess A, Barberá P, Vaccari C, Siegel A, Sanovich S, Stukal D, Nyhan B (2018) Social media, political polarization, and political disinformation: A review of the scientific literature. SSRN Electron J. https://doi.org/10.2139/ssrn.3144139 Vaccari C, Chadwick A (2020) Deepfakes and disinformation: Exploring the impact of synthetic political video on deception, uncertainty, and trust in news. Social Media + Soc. https://doi.org/10.1177/2056305120903408 van Dijck J, Poell T, de Waal M (2018) The platform society: Public values in a connective world. Markets Globalization Dev Rev 3(3). Article 6. https://doi.org/10.23860/MGDR-2018-03-03-08 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 Wardle C, Derakhshan H (2017) Information disorder: Toward an interdisciplinary framework for research and policy making (DGI(2017)09). Council of Europe. https://rm.coe.int/information-disorder-toward-an-interdisciplinary-framework-for-researc/168076277c Woolley SC (2020) The reality game: How the next wave of technology will break the truth, and what we can do about it . Endeavour. https://www.worldcat.org/oclc/1127085781 Woolley SC, Howard PN (2017) Computational propaganda worldwide: Executive summary (Working Paper No. 2017.11). University of Oxford Zannettou S, Sirivianos M, Blackburn J, Kourtellis N (2019) The web of false information. In Proceedings of the ACM Internet Measurement Conference . https://doi.org/10.1145/3309699 Zhou X, Zafarani R (2021) A survey of fake news: Fundamental theories, detection methods, and opportunities. ACM-CSUR 53(5):1–40. https://doi.org/10.1145/3395046 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 07 May, 2026 Submission checks completed at journal 06 May, 2026 First submitted to journal 06 May, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9473319","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":636169549,"identity":"82404f94-5a29-4235-b12f-1ac58a2bd7e6","order_by":0,"name":"Nigar Garajamirli","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABEklEQVRIie2RMUsDMRTHcxzc9LCbXEm5fALhlQMXoX4L54SCWVro6CBypdDR+SCDX8EiOAcC+iG6VPwCmcRB1Hep4NI7OwrNDxJe4P3yzyOMRSL/kqRq9kFG1UbiV9Gc7GYPBUhJ0c/SMihyjyyglfVrn6pwQ5fSO3ZLz64dHPGKccBM3124F0oZFSfVbqVv1KJmTw6ygWUlIExX60skZVye2t0KrtWCnkRKTm2A+XRlZKNY9diinAflc6s4QNRDo32ngpyUZBmUZF6jlIJPulPyJkXdalKo8GiH93wysxLbZ+kZ/cr821kh6vHzu/ywQhj94P3VqGhTApLRPD9/yjB0Ykf7lpvfUlR/dkcikciB8Q2GjVv7t8J/uQAAAABJRU5ErkJggg==","orcid":"","institution":"","correspondingAuthor":true,"prefix":"","firstName":"Nigar","middleName":"","lastName":"Garajamirli","suffix":""}],"badges":[],"createdAt":"2026-04-20 14:08:32","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9473319/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9473319/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108940422,"identity":"0d22dac8-b57d-4008-bde5-1e303e709e03","added_by":"auto","created_at":"2026-05-11 05:12:44","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":90540,"visible":true,"origin":"","legend":"\u003cp\u003eAlgorithmic Amplification and Synthetic Influence Framework\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/0863b3085e0620b2c744e0e1.png"},{"id":108940365,"identity":"bc41196b-88f7-4416-8018-65e9f244668f","added_by":"auto","created_at":"2026-05-11 05:12:22","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1423580,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThree-Dimensional Distribution of User Tweet Activity\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/ec11c155a314587b778bfa3e.png"},{"id":108940380,"identity":"4da4e407-7e58-4150-a921-a071939f41a1","added_by":"auto","created_at":"2026-05-11 05:12:32","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":140830,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRetweet Amplification Landscape\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/dc2a21ccddf23c9fa1aa5189.png"},{"id":108940415,"identity":"a97558f6-47e3-48db-a3db-e4e67a6d48b2","added_by":"auto","created_at":"2026-05-11 05:12:41","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":257950,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBehavioral Space of User Activity\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/ee22fe9a2fae337353403f8b.png"},{"id":108940374,"identity":"6398edef-b15a-4ba2-9143-462376bbc5f8","added_by":"auto","created_at":"2026-05-11 05:12:26","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":121386,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBehavioral Threshold Detection Space\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/d31d6e9b2e1340a3699f7cdf.png"},{"id":108940392,"identity":"5afa958a-0fae-43c5-a6e9-429536178724","added_by":"auto","created_at":"2026-05-11 05:12:38","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":287029,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eActivity and Amplification Correlation Landscape\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/b8a5e7037b5bf1ee6b655295.png"},{"id":108940396,"identity":"a9c15781-c1dc-4d4a-9868-b2e94c4ee139","added_by":"auto","created_at":"2026-05-11 05:12:40","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":259446,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBot-Driven Information Diffusion\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/6abd62b049d5c8c402f6530d.png"},{"id":108940414,"identity":"3bedbac2-85a7-4105-a9a0-b58236713df2","added_by":"auto","created_at":"2026-05-11 05:12:40","extension":"jpeg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":106198,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSynthetic Political Communication Network with Bot Amplification\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/d0afad3c2557082db204d296.jpeg"},{"id":108940394,"identity":"27ccacf3-5672-45a4-a91e-84f5f9e263c8","added_by":"auto","created_at":"2026-05-11 05:12:40","extension":"jpeg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":65248,"visible":true,"origin":"","legend":"\u003cp\u003eLog–log degree distribution\u003c/p\u003e","description":"","filename":"image9.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/60507cf22a139e987a6b3e37.jpeg"},{"id":108940382,"identity":"32116d7b-12d6-4d01-9fc5-08d1a171ff54","added_by":"auto","created_at":"2026-05-11 05:12:33","extension":"jpeg","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":150940,"visible":true,"origin":"","legend":"\u003cp\u003eCentrality amplification due to bots\u003c/p\u003e","description":"","filename":"image10.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/e0f803344c7426ef616638c0.jpeg"},{"id":108940421,"identity":"d3e1eda4-436a-4316-89b6-616bb117d632","added_by":"auto","created_at":"2026-05-11 05:12:44","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":96583,"visible":true,"origin":"","legend":"\u003cp\u003eComparative Communication Dynamics (2016 vs 2024)\u003c/p\u003e","description":"","filename":"image11.png","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/a90913dfa0b49aea2dd511ed.png"},{"id":108940416,"identity":"01ea5712-9309-49ad-9e5a-a4b576d3453e","added_by":"auto","created_at":"2026-05-11 05:12:41","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":217110,"visible":true,"origin":"","legend":"\u003cp\u003eInfluence Density Landscape\u003c/p\u003e","description":"","filename":"image12.png","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/f77c71598400a83ca47eda62.png"},{"id":108940417,"identity":"a8823e5b-a3c4-46e4-8f34-89fd86342bc2","added_by":"auto","created_at":"2026-05-11 05:12:41","extension":"png","order_by":13,"title":"Figure 13","display":"","copyAsset":false,"role":"figure","size":248736,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGalaxy Map of Political Communication Networks\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image13.png","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/72c04a271fd3237cf5eb4257.png"},{"id":108940426,"identity":"fd08d1d7-e0be-4d5c-853c-42c4bfc8ddb0","added_by":"auto","created_at":"2026-05-11 05:12:51","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3517151,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9473319/v1/a773e727-1b0b-438c-89f3-de89a30a340e.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Truth, Trust, and Technology: Political Communication in the Era of Synthetic Media","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe fast spread of the digital communication technologies has led to certain radical changes in the ways how we communicate politically and tend to participate in democratic life. Today, social media platforms play a role of the central forums for political debatштп, which enable politicians, journalists, activists, and citizens to be able to interact within a scope of highly interconnected media environment. These platforms can facilitate the fast enough dissemination of information, real-time political engagement and the mobilization of public opinion on a pretty large scale, which fundamentally alters the means how political discourse is shaped and in result disseminated in modern society (Tucker et al., \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Kreiss, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Jungherr, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Schillemans, \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Notably in this digital ecosystem, communications are seen to be increasingly mediated by algorithmic infrastructure, which in turn organizes, categorizes, and distributes political content across the entire network (Napoli, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Gillespie, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Helberger et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Bucher, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Beer, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eA key point is that, with the increasing influence of the infrastructure of platforms and algorithmic systems new opportunities are notably emerging in the realms of political movements and civic engagement. As a consequence, politicians are believed to increasingly rely on strategies with the data being in the core, personalized messaging and targeted advertising to be able to reach voters online (Aral \u0026amp; Eckles, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Buts, 2021; Kreiss et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). As a result, digital platforms tend to become deeply integrated into the process of the modern election, which in turn lead to fundamental changes in the strategies of the traditional election and expanding the role of online debate regarding the political discourse altogether shaping public opinion (Karpf, 2024; Bimber, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOn the other hand, the technological infrastructure that makes possible broad political participation tend to create new vulnerabilities in the communication systems of the democratic world. A plenty of studies indicate that the fast enough spread of misinformation, the campaigns of the organized disinformation alongside with the emergence of artificial media may pose truly big challenges to modern democracy (Lazer et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Wardle \u0026amp; Derakhshan, \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Farkas \u0026amp; Schou, 2020; Benkler et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). The misinformation and falsehoods are believed to be spread rapidly on social media, which in turn often reaching many people, sometimes even faster than the verified information (Vosoughi et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Therefore, the exposure to content of this type can be often limited to specifically defined ideological communities, resulting in reinforcing the existing belief systems among the people and exacerbate the polarization among politics (Guess et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Grinberg et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Cinelli et al., 2020).\u003c/p\u003e \u003cp\u003eIt is worth noting that, the recent up-to-date advances in the area of the artificial intelligence and generative technologies have made these challenges worse. As a consequence, AI-generated texts, images, and videos (including \"deepfakes\") tend to create the new forms of artificial communication in politics which is capable of manipulating political discourse and influencing public opinion (Battista, 2024; Kohanets, 2025; Bommasani et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Floridi et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Therefore, these technologies make the automated production of massive amounts of interactive content possible, which in result make it pretty difficult for users to find the difference between the original and automated information. In some possible cases, the engagement with AI-generated content was close to equal or even it was possible to surpass engagement with the content generated with human, which could raise concerns about the potential for possible manipulation of the environment of digital information.\u003c/p\u003e \u003cp\u003eOne of the most well-known and significant mechanisms in terms of artificial influence in online political communication is the utilization of of automated social media accounts, which are widely known as \"social bots.\" These bots are believed to be algorithmically controlled accounts, which are capable of generating content highly independently, interacting with other users alongside with other functions like amplifying political messages within digital networks (Ferrara et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Howard \u0026amp; Kollanyi, 2016; Ferrara, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn accordance with the experimental studies, organized botnets can somehow be the reason of distorting political discourse via generating massive amounts of messages and at the same time artificially amplify certain narratives (Bessi \u0026amp; Ferrara, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Shao et al., \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Subrahmanian et al., \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Moreover, organized bot activity is tended to be observed during events regarding political activities like national elections, where it strongly manipulates the flow of information and could shape online discussions via using the strategies commonly recognized as computational propaganda (Badawy et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Bradshaw \u0026amp; Howard, 2018; Woolley \u0026amp; Howard, \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe possible presence of automated actors within social media platforms can be considered as a significant threat to the communication within democratic society. These automated programs can amplify unreliable sources, reinforce closed ideological affiliations, and spread misleading narratives within divided societies (Cinelli et al., 2020; Bovet \u0026amp; Makse, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Stella et al., \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). During election campaigns, organized disinformation activities are tend to be increasingly documented, taking into account the automated accounts, which strategically spread political claims and misinformation in order to influence public opinion and behavior of the voter (Cinus et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Im et al., 2020).\u003c/p\u003e \u003cp\u003eDespite the fact, that the research regarding the disinformation via computer generation and propaganda could make great strides, a significant gap still remains nowadays in the empirical understanding in terms of the dynamics of artificial political communication (Das \u0026amp; Clark, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). To be precise, a deeper understanding of the ways how automated actors tend to influence political discourse in the environments of the social media requires the integrated approach based on computations that can simultaneously analyse the patterns of the engagement, network structures altogether with amplification dynamics. Even though previous research had a focus of analysing bot detection and disinformation dissemination separately, relatively few studies are believed to integrate these analytical perspectives into a unified framework of the computations (Ferrara, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Shao et al., \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Cresci et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWith the aim of the bridging this gap, this study conducts a series of computational analyses in order to examine patterns of communication within the political environment and artificial amplification within social networks. Via extensive Twitter data on election-related discussions, author analyses the distribution of all user activity, the dynamics of retweet amplification, metrics of bot-like behavior, finishing with the the network architecture that can potentially shape information dissemination. Furthermore, author could employ modelling within statistical frames and network metric analysis with the aim to assess how user engagement and message amplification can contribute to influence within the area of social networks and the ways how bot actors can possibly alter the overall dynamics of dissemination.\u003c/p\u003e \u003cp\u003eThis research is be conducted within the broader framework of communication of the digital political world and computational social science. Recent research in this field could reveal the ways how social media platforms tend to reshape political discourse via mass engagement, algorithmic authorization mechanisms and the disseminationof the information based on the network. In such digitally mediated environments, the political communication is increasingly appearing from certain types of interactions between human users, platform algorithms and automated agents. This study aims to contribute to a truly deeper understanding of the certain ways how the structural characteristics of communication networks can possibly influence the flow and visibility of information within the political context in modern digital public spaces via applying the computational methods to large datasets within the social media environment.\u003c/p\u003e \u003cp\u003eResearch Contributions\u003c/p\u003e \u003cp\u003eThis study makes highly valuable contributions to the fields of digital political communication and computational sociology.\u003c/p\u003e \u003cp\u003eMore specifically, it could conduct a comparative analysis within the computational environment of political communication across two majorly known election cycles, which in result revealed measurable structural changes in the participation of the variability, influence intensity alongside with the network centralization.\u003c/p\u003e \u003cp\u003eSecond, it could provide a comprehensive analytical framework that enabled to systematically link user activity, message amplification and influence dynamics, which makes a\u003c/p\u003e \u003cp\u003emultidimensional assessment of communication processes within the frames of the digital environment possible.\u003c/p\u003e \u003cp\u003eThird, it could quantify the amplification effect of accounts considered highly active and potentially automated, which could potentially demonstrate how a presumably tiny number of actors can disproportionately influence the process of wide the information dissemination.\u003c/p\u003e \u003cp\u003eFinally, there is a high contribution is to sociology within the computational measurements via demonstrating the potential in order to reveal the underlying structural characteristics of systems regarding the political communication through the ways of integration of large-scale digital tracking data with the feature of simulation modeling.\u003c/p\u003e \u003cp\u003eIt should be mentioned that this study could surpass previous research in terms of quantitatively monitoring structural changes in the realm of the digital political communication between the period of the indicated election cycles. This study, in particular, indicates that a relatively tiny number of highly active and considered as potentially automated computations can measurably make the information reconstructed in terms of diffusion patterns and network centralization. Using the commonly known ways like integrating behavioral indicators, amplification dynamics and the architectures within network system into a unified computational framework, the current research uses a systematic approach in order to analyze the complex effects in communication environments based on the Twitter platform.\u003c/p\u003e \u003cp\u003eResearch Questions\u003c/p\u003e \u003cp\u003eBased on the theoretical framework and research objectives mentioned above, this study mainly tends to address the following research questions:\u003c/p\u003e \u003cp\u003eRQ1: How is political communication distributed across users in election-related discussions on social media, and how do participation patterns shape the production of political content?\u003c/p\u003e \u003cp\u003eRQ2:\u003c/p\u003e \u003cp\u003eHow do amplification mechanisms in digital networks\u0026mdash;including retweet dynamics, user network positions, and automated actors\u0026mdash;affect the dissemination and visibility of political messages?\u003c/p\u003e \u003cp\u003eHypotheses\u003c/p\u003e \u003cp\u003eTo empirically test these research questions, the study examines the following hypotheses.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eHypothesis 1\u003c/strong\u003e \u003cp\u003ePolitical activity on social media is highly not evenly distributed, with only a small number of users who produce the majority of the political content.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eHypothesis 2\u003c/strong\u003e \u003cp\u003eMessage amplification through retweets can lead to a highly biased dissemination pattern, where a very small percentage of messages can achieve significantly wider reach comparing to the majority of content.\u003c/p\u003e \u003c/p\u003e \u003cp\u003eThird hypothesis: Accounts that behave automatically exhibit significantly higher levels of activity comparing to the ordinary human users.\u003c/p\u003e \u003cp\u003eFourth hypothesis: Users who take a central position in the social network are inclined to exert greater influence on the dissemination of political information.\u003c/p\u003e \u003cp\u003eFifth hypothesis: The presence of automated bots in the social network can generally increase the potential spread and amplification of political messages.\u003c/p\u003e"},{"header":"2. Literature Review and Theoretical Framework","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003e2.1. Digital Political Communication in Platform Societies\u003c/h2\u003e\n \u003cp\u003eThe transformation within the political communication system in the digital age can be closely linked to the sudden emergence of a communication infrastructure, which is based on digital platforms. Social media platforms like Twitter, Facebook and YouTube are inclined to tuning to into the main centres for the political exchange, which in result let politicians, journalists, activists, and citizens share their information and actively take place in public discourse. Researchers are inclined toward increasingly viewing this transformation via the lens of the platform society, where the production, distribution, and dissemination of information within modern communication systems is actively shaped by digital infrastructure (van Dijck et al., \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Poell et al., 2019; Gorwa et al., 2020; Kitchin, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eTaking into account these details, political communication is increasingly considered as a subject to algorithmic systems in which the interaction process and visibility are prioritized and the spread of information online is considered as common. Platform algorithms can currently influence which political messages are visible, how information has been circulating online and which narratives dominate public discourse (Gillespie, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Napoli, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Bucher, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Beer, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Therefore, social media platforms are not merely considered as communication tools as they have turned to be structural intermediaries which are able to shape the dynamics of information flow within the politics.\u003c/p\u003e\n \u003cp\u003eThe sudden rise of platform communication could fundamentally alter the democratic inclusion and political campaigning. Politicians currently are more likely to rely heavily on digital platforms in order for them to deliver their messages, mobilize supporters and engage directly with the voters. Communication strategies driven on the data, targeted political advertising and campaign tactics that are algorithm-driven are currently playing an increasingly major role in contemporary electoral politics (Kreiss, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Karpf, 2024; Bimber, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Vaccari, 2013).\u003c/p\u003e\n \u003cp\u003eThis shift has been further envisioned via the theory of networked action, which emphasizes the ways how digital network communications enable decentralized forms of political mobilization that is operating outside traditional organizational structures (Bennett \u0026amp; Segerberg, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Loader \u0026amp; Mercea, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Theocharis et al., \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). In contrast to the traditional model of political communication, where mass media served as the primary gateways to political information (Habermas, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), the digital environment has enabled individuals to simultaneously create, distribute and then to disseminate political content.\u003c/p\u003e\n \u003cp\u003eThe well-known algorithmic frameworks of the platforms for social media, however, is presenting reasonable obstacles to democratic discourse. Via providing more weight to information that in result can fit with the beliefs of users, the recommendation systems fully personalized for users can potentially make ideological divides worse and in result create echo chambers and highly selective patterns in the ways how the chunks of the information is presented (Pariser, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Sunstein, \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Cinelli et al., 2020). Empirical research in this case demonstrates that these dynamics can make political polarization even worse and change drastically typical patterns of political engagement in the digital realm (Iyengar \u0026amp; Hahn, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Stroud, \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e2011\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eSimultaneously, the extensive digital ramifications coming from online interactions present the straight and clear opportunities to be able to analyze communication patterns, ideological networks and the ways of information is disseminated. As a result, social science methods based on the computations are becoming more and more important for studying how digital political communication works (Jungherr et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Barber\u0026aacute; et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2015\u003c/span\u003e).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n \u003ch2\u003e2.2. Misinformation and Information Disorder\u003c/h2\u003e\n \u003cp\u003eThe wide spread of misinformation is considered as a central issue in digital communication of the modern world. Information disorder can be commonly divided into 3 categories: misinformation (false information shared without the intent to deceive), disinformation (deliberately misleading content) and malinformation (pretty accurate information used with the aim to cause harm) (Wardle \u0026amp; Derakhshan, \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Lazer et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Allcott \u0026amp; Gentzkow, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eEmpirical studies showed that misinformation spreads fast across social media, often spreading even faster than expected and reaching larger audiences than the information identified as the accurate one (Fusoggi et al., 2018). Its circulation is seen to be frequently concentrated within the frames of the ideologically aligned communities that constantly tend to reinforce shared narratives (Grinberg et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Guess et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eThe spread of misinformation is also seen and influenced by known psychological factors like cognitive biases, emotional responses, identity-based reasoning and partisan motivations (Lewandowsky et al., 2017; Pennycook \u0026amp; Rand, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; van der Linden et al., 2020). Simple interventions, the one is especially known as the prompting users to consider the accuracy before sharing, have been shown to significantly reduce the amount or as a best outcome avoid misinformation sharing (Pennycook et al., 2020).\u003c/p\u003e\n \u003cp\u003eAt the structural level, algorithms of the known platform and media systems based on the attention sharing tend to amplify emotionally engaging or the content considerd more sensational, inadvertently increasing the possible visibility of misleading narratives (Benkler et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Napoli, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). As a result, these dynamics may collectively contribute to a certain environment where the misinformation gains the high traction and can as an outcome influence public opinion.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n \u003ch2\u003e2.3. Computational Propaganda and Automated Actors\u003c/h2\u003e\n \u003cp\u003eComputational propaganda is considered to be bonded with the wide utilization of automated systems, algorithms and coordinated networks with the final aim to manipulating the public opinion online (Woolley \u0026amp; Howard, \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Bradshaw \u0026amp; Howard, 2018).\u003c/p\u003e\n \u003cp\u003eA core element of these advanced strategies is involving social bot, which are defined as the automated accounts capable of generating content and interacting with users in order to amplify messages, spread disinformation and in result to increase visibility (Ferrara et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Howard \u0026amp; Kollanyi, 2016).\u003c/p\u003e\n \u003cp\u003eA plenty of empirical researches have demonstrated that automated accounts are set to actively participating in political discussions and contribute to the pretty active dissemination of unreliable information, particularly taking into account the activity on the platforms like Twitter (X) (Shao et al., \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Subrahmanian et al., \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). These accounts have the potential to distort information flows via the ways like rapidly reposting content, simulating engagement and coordinating messaging across networks (Stella et al., \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eAs a result, detecting automated behavior has become a key focus in computational social science, with the methods of the machine learning developed to identify bots based on behavioral, linguistic and features which are based on the network (Cresci et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Zhou \u0026amp; Zafarani, 2020).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\n \u003ch2\u003e2.4. Artificial Intelligence and Synthetic Political Communication\u003c/h2\u003e\n \u003cp\u003eModern advancements in the realm of artificial intelligence have led to the introduction of the synthetic media as a new dimension of political communication in the digital world. Generative AI systems currrently have the abilities to producing text, images, audio and video which tend to closely resemble content created by human.\u003c/p\u003e\n \u003cp\u003eA particularly concerning issue nowadays is seen in the development is deepfake technology, which allows the creation of highly realistic and at the same time artificially generated material with the sceptrum of audiovisual info. Content of that ilk shows high risks to democratic discourse via the ways like facilitating the spread of convincing but at the same time the misleading messages in political realm (Vakari \u0026amp; Chatwick, 2020).\u003c/p\u003e\n \u003cp\u003eStudies show that the content generated via AI usage can achieve levels of engagement which can be compared to or exceeding human-generated content, which in result can increas its potential influence. As generative AI becomes more common, the concerns regarding its currrent role in shaping political narratives and public opinion is still seen to actively grow.\u003c/p\u003e\n \u003cp\u003eAt the same time, the common detection methods are tend to be developed using machine learning techniques that have the set to analyze visual inconsistencies, behavioral signals and dissemination patterns with the aim to identifing manipulated synthetic media.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\n \u003ch2\u003e2.5. Algorithmic Governance and Democratic Implications\u003c/h2\u003e\n \u003cp\u003eThe increasing role of systems with the certain algorithms in digital platforms has led to important questions about the possible accountability in terms of governance and democratic. Platform algorithms influence the ways which political content users see, how information can circulate and which narratives currently seen to dominating public discourse (Helberger et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Kitchin, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eThese systems can often be described as opaque \u0026ldquo;black boxes,\u0026rdquo; as it is inclined to limit transparency and make it difficult for users, regulators altogether with the famous researchers with the aim to understanding how information is filtered and distributed (Pasquale, 2015; O\u0026apos;Neil, 2016).\u003c/p\u003e\n \u003cp\u003eThis issue can be observed to be closely connected to broader debates on digital political economy, data power and capitalism with the focus on surveillance (Couldry \u0026amp; Mejias, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Fuchs, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Taking into account the current growing integration of AI into certain communication infrastructures, concerns around transparency, accountability altogether with governance have increasingly seen as a core in contemporary research (Floridi et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Mittelstadt et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003ch2\u003e2.6. Algorithmic Amplification and Synthetic Influence Framework\u003c/h2\u003e\n \u003cp\u003eBased on ideas from platform research, disinformation research and computational propaganda research, the study provides an \u0026quot;algorithmic amplification and artificial influence framework\u0026quot; in order to conceptualize the ways how political messages can gain recognition and influence within systems of digital communication.\u003c/p\u003e\n \u003cp\u003eThe Amplification Framework in Digital Political Communication\u003c/p\u003e\n \u003cp\u003eThis framework shows the dissemination of information in political realms as a multi-stage amplification process, which can be driven by three interacting mechanisms:\u003c/p\u003e\n \u003cul\u003e\n \u003cli\u003e\n \u003cp\u003eUnequal Participation: Political content production can be seen as highly concentrated, with a tiny number of users producing the majority of the current content. This in result reflects the pattern of the power distribution that characterizes networks in the digital world.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eAlgorithmic Amplification: Platform algorithms are inclined to prioritize content related to the high-engagement process (likes, shares, retweets, etc.) and boost its possible visibility through a feedback loop which is actively seen in already popular messages.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eArtificial Amplification: Automated or semi-automated entities like the social media bots which amplify these dynamics via the ways that generate massive amounts of content, promote specific messages and inflate engagement metrics.\u003c/p\u003e\n \u003c/li\u003e\n \u003c/ul\u003e\n \u003cp\u003eThe convergence of these highly clever mechanisms creates a communicative environment in which a tiny and not that noteble number of entities can exert a disproportionate influence. Therefore, online political discourse can possibly arise from the interaction between human users, algorithmic system and automated entities.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\n \u003ch2\u003e2.7. Research Gap and Study Contribution\u003c/h2\u003e\n \u003cp\u003eCurrent research seem to indicate that political communication can potentially be influenced by digital platforms, algorithms, automated actors alongside with the techniques related to the artificial intelligence. However, few studies could have addressed these dynamics in the environments of the artificial media via the ways of combining specially selected large-scale election data with the sophisticated analysis of the computational network and behavioral modeling.\u003c/p\u003e\n \u003cp\u003eThis study aims to fill this gap via the broad analysis of the political communication and information dissemination in discussions regarding the elections on Twitter (X) using computational social science methods. Via integrating statistical analysis, network modeling, and simulation, author provides empirical insights into the ways of how platform is structured and automated actors can possiblt affect the spread of political discourse.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003e3.1. Research Design\u003c/h2\u003e\n \u003cp\u003eThis study seems to adopt a framework of the computational social science with the aim to analyze the structural dynamics within the frames of political communication and artificial amplification on the social media. The research design currentlycombines data analysis with the large-scale dataset, behavioral modeling, network simulations and statistical methods aiming at examining user activity, message amplification and also information diffusion.\u003c/p\u003e\n \u003cp\u003eVia the utilization of large-scale digital trace data taken with the computational modeling allows for investigation on the systematic basic of patterns related to thecommunication that are difficult to capture thorugh using traditional qualitative or small-scale experimental methods.\u003c/p\u003e\n \u003cp\u003eThe currently given empirical analysis focuses on Twitter (X), a widely used and well-known platform in studies related to the political communication, disinformation and automated behavior. Its publicly available data like tweets, retweets, replies and engagement metrics making the analysis prcoess in the large scales of communication networks and discourse possible.\u003c/p\u003e\n \u003cp\u003eThe study consists of a plenty of computational experiments, which were conducted in MATLAB, each specifically targeting a specific structural aspect of political communication, including:\u003c/p\u003e\n \u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n \u003cul\u003e\n \u003cli\u003e\n \u003cp\u003eThe distribution of user engagement\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eThe retweet-driven amplification dynamics\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eThe behavioral differences between typical and highly active users\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eThe detection of bot-like activity patterns\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eThe relationships between engagement metrics\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eThe network propagation under the automated amplification\u003c/p\u003e\n \u003c/li\u003e\n \u003c/ul\u003e\n \u003cp\u003eTogether, these components can form a sophisitcated framework with multiple layers for understanding the methods of how political the information spreading and how automated actors can possibly shape the dynamics of the communication in the environments of the digital world.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003e3.2. Dataset Construction and Data Sources\u003c/h2\u003e\n \u003cp\u003eThe given empirical analysis is based on a truly massive dataset of tweets related to the election process posted on Twitter (X) during the 2016 and 2024 US presidential elections. The dataset takes into account data of both tweet-level and user-levels, which allows the analysis of engagement patterns, message dissemination dynamics and the characteristics related to the user behaviour in discussions on political topics.\u003c/p\u003e\n \u003cp\u003eEach of the given observation in the dataset includes multiple variables, which describe communication behavior, such as:\u003c/p\u003e\n \u003cp\u003eTweet content\u003c/p\u003e\n \u003cp\u003eTimestamp\u003c/p\u003e\n \u003cp\u003eUser ID (anonymous)\u003c/p\u003e\n \u003cp\u003eNumber of retweets\u003c/p\u003e\n \u003cp\u003eEngagement metrics\u003c/p\u003e\n \u003cp\u003eHashtag usage patterns\u003c/p\u003e\n \u003cp\u003eThe integrated and collected dataset includes approximately 750,000 tweets created by more than the 80,000 users. Specifically, the final dataset is consistsing of:\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e355,000 tweets from the 2016 election period\u003c/p\u003e\n \u003cp\u003e402,000 tweets from the 2024 election period\u003c/p\u003e\n\u003c/div\u003e\n\u003cp\u003eThese datasets make possible a comparative study of political communication dynamics during the two election periods.\u003c/p\u003e\n\u003cp\u003eThe datasets which are analyzed in this study were obtained from a widely-known and publicly available Twitter dataset on the so-called Cagle Research Data Platform. These datasets currently containing tweets with the main topic of election which is collected via the Twitter API and shared for research purposes. They at the same time include tweet-level metadata alike as timestamps, engagement metrics, anonymous user IDs and retweet counts, making a comprehensive analysis of dynamics within the political communication realm in the digital political environment feasible\u003c/p\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003ch2\u003e3.3. Data Preprocessing\u003c/h2\u003e\n \u003cp\u003ePrior to the current analysis, the chosen dataset went though several preprocessing procedures in order to guarantee the consistency and reliability in terms of analysis. The preprocessing pipeline included:\u003c/p\u003e\n \u003cp\u003eThe removal of duplicate tweets\u003c/p\u003e\n \u003cp\u003eThe filtering of incomplete or deleted records\u003c/p\u003e\n \u003cp\u003eThe normalization of timestamp formats\u003c/p\u003e\n \u003cp\u003eThe verification of temporal consistency across observations\u003c/p\u003e\n \u003cp\u003eThe extraction of engagement indicators such as retweet counts\u003c/p\u003e\n \u003cp\u003eUser-level activity metrics were subsequently constructed with the aim to measure the participation intensity across the communication network.\u003c/p\u003e\n \u003cp\u003ePreviously analysed researches have demonstrated that the participation within the environments of the social media typically follows heavy distributions, in which a relatively not noticeable number of users can lead to the generation of a disproportionately large share of content. Preliminary inspection of the previously chosen dataset confirmed that patterns of the similar participation are present in the empirical data are analysed in this study.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003e3.4. Behavioral Modeling of Communication Activity\u003c/h2\u003e\n \u003cp\u003eThe behavior of the users was analyzed via utilization of metrics based on the activity, which is derived from the dataset, taking into account the number of tweets per user, retweet rate, engagement intensity and the index of a composite activity. The idea of the index is to combine posting frequency and retweet activity with the goal to capture overall communication intensity in the discourse of the politics.\u003c/p\u003e\n \u003cp\u003eChosen users above the 95th percentile of activity were inclnied to be classified as highly active. This classification can not directly identify bots but at the same time it follows a common approach in behavioral studies within the computational social science where bot detection tools are not that widely available. Such accounts with the high activity may include automated agents, coordinated accounts or even human users with presumably high engagement.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n \u003ch2\u003e3.5. Synthetic Network Construction\u003c/h2\u003e\n \u003cp\u003eTaking into acount the goals of examining self-amplification effects, a network of the artificial communication was constructed with the full implementation of a scale-free (non-scale) model, which is based on preferential attachment. This approach can reflect social media structures seen and identified in the real-world, in which highly connected users tend to attract more interactions.\u003c/p\u003e\n \u003cp\u003eThe network was in result initialized with 3 fully connected nodes, with additional nodes added step by step based on the preferential attachment (i.e., connection probability proportional to node connectivity). The initial network of the experiment included 600 nodes taken to analyze structural properties in a controlled environment.\u003c/p\u003e\n \u003cp\u003eLarger simulations were conducted with 5,000 nodes created to test robustness, particularly with the aim of analyzing automated amplification dynamics (see Section \u003cspan refid=\"Sec35\" class=\"InternalRef\"\u003e4.7\u003c/span\u003e). Even though it is smaller than real-world networks, these simulations make key structural features viisble and give approximate meanings and provide a controlled framework with studying the effects of self-amplification in the systems of the digital communication.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\n \u003ch2\u003e3.6. Modelling Automated Amplification\u003c/h2\u003e\n \u003cp\u003eTo investigate the possible impact of actors with full automatization, bots were introduced into an artificial network.\u003c/p\u003e\n \u003cp\u003eTh conducted by author simulation compares two experimental scenarios:\u003c/p\u003e\n \u003cp\u003eThe base network (consisting only of human users)\u003c/p\u003e\n \u003cp\u003eThe enhanced network (with the addition of automated bot accounts)\u003c/p\u003e\n \u003cp\u003eIn the enhanced scenario, automated actions were actually distributed within approximately 400 nodes, which are based on the previously mentioned rating showing the high activity.\u003c/p\u003e\n \u003cp\u003eThe bots were designed as nodes showing high-activity that actively created and retweeted content and also mirroring the actual activity patterns that are observed in the dataset.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\n \u003ch2\u003e3.7. Bot Amplification Mechanism\u003c/h2\u003e\n \u003cp\u003eAuthor has simulated an auto-amplification strategy with the way of allowing bot nodes in order to create additional links beyond the underlying network structure.\u003c/p\u003e\n \u003cp\u003eFor each bot node:\u003c/p\u003e\n \u003cp\u003eBetween 5 and 10 additional links were created.\u003c/p\u003e\n \u003cp\u003e70% of the new links targeted higher-order nodes (network hubs).\u003c/p\u003e\n \u003cp\u003e30% of the links targeted randomly selected users.\u003c/p\u003e\n\u003c/div\u003e\n\u003cp\u003eThis mechanism could successfully mimic the strategies of auto-amplification, which are common in online political networks, in which automated accounts can increase the probable visibility of messages with the way of frequently interacting with influential users and trending topics.\u003c/p\u003e\n\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e\n \u003ch2\u003e3.8. Network Metrics\u003c/h2\u003e\n \u003cp\u003eIn order to assess the structural differences between organic and amplified communication networks, author could calculat several network metrics:\u003c/p\u003e\n \u003cp\u003eMean score: It shows the average number of connections per node.\u003c/p\u003e\n \u003cp\u003eExtreme score: It indicates the influence of necessary nodes with a plenty of connections within the network.\u003c/p\u003e\n \u003cp\u003eAggregation coefficient: This indicates the tendency of nodes with the goal to form closely interconnected local clusters.\u003c/p\u003e\n \u003cp\u003eFurthermore, author also investigated the following distributions:\u003c/p\u003e\n \u003cp\u003eScore distribution\u003c/p\u003e\n \u003cp\u003eScore centrality distribution\u003c/p\u003e\n \u003cp\u003eVia utilization of these metrics, author was able to conduct a structural comparison between organic and communication networks with the self-amplification.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec22\" class=\"Section2\"\u003e\n \u003ch2\u003e3.9. Correlation Analysis\u003c/h2\u003e\n \u003cp\u003eWith the aim to analyze the relationships between communication measures, a correlation analysis was conducted on multiple interaction variables.\u003c/p\u003e\n \u003cp\u003eThe variables analyzed included:\u003c/p\u003e\n \u003cp\u003eTweet frequency\u003c/p\u003e\n \u003cp\u003eRetweet count\u003c/p\u003e\n \u003cp\u003eInteraction intensity\u003c/p\u003e\n \u003cp\u003eAn impact index calculated from the interaction measures.\u003c/p\u003e\n \u003cp\u003ePearson\u0026apos;s correlation coefficient was implemented in order to assess the statistical relationships between the chosen variables.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec23\" class=\"Section2\"\u003e\n \u003ch2\u003e3.10. Experimental Implementation\u003c/h2\u003e\n \u003cp\u003eAll given computational experiments were conducted through the MATLAB program, which provides a big number of tools for numerical calculations, statistical modelling and graphics.\u003c/p\u003e\n \u003cp\u003eThe experimental framework utilized the following tools:\u003c/p\u003e\n \u003cp\u003eStatistical analysis functions\u003c/p\u003e\n \u003cp\u003eNumerical modeling tools for simulating behavior\u003c/p\u003e\n \u003cp\u003eGraphalographic tools for representing multidimensional data\u003c/p\u003e\n \u003cp\u003eThe MATLAB scripts could generate the graphs illustrating the structural patterns of communication within the political envirnoment, including activity distribution, retweet amplification dynamics and the processes of the simulated network propagation.\u003c/p\u003e\n \u003cp\u003eAll provided graphs were exported at 600 dots per inch (DPI) to meet the standards of academic publication.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec24\" class=\"Section2\"\u003e\n \u003ch2\u003e3.11. Statistical Significance\u003c/h2\u003e\n \u003cp\u003eThe statistical significance of the observed differences was evaluated via using the method of the standard inference. All statistical tests were made at a 95% confidence level and following results with a p-value less than 0.05 could be considered statistically significant.\u003c/p\u003e\n \u003cp\u003eAdditional tests were conducted with the aim to verify the robustness of the results, making possible the fact that the observed communication patterns were not influenced by individual observations or any other factors. The consistency of the relationships with certain structures in the two election datasets could support the stability of the experimental results.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec25\" class=\"Section2\"\u003e\n \u003ch2\u003e3.12. Methodological Limitations\u003c/h2\u003e\n \u003cp\u003eFirst thing to be noted, the analysis relies on Twitter (X) data, which can represent only a bordered segment of the broader landscape into the political communication. Political discourse may also occur through other channels, taking into account traditional media, private messaging platforms and interactions made offline that are not captured in public datasets.\u003c/p\u003e\n \u003cp\u003eSecond, the identifying automated or coordinated accounts which is based solely on behavioral metrics is considerd as inherently difficult. High activity levels may indicate automation, but at the same time they can also possibly reflect highly engaged human users.\u003c/p\u003e\n \u003cp\u003eAdditionally, social media platforms frequently tend to changing their algorithms and policies of the moderation, which can in result influence communication patterns and information dissemination over specified period of time.\u003c/p\u003e\n \u003cp\u003eGiven these constraints, the findings should be interpreted with caution. The applied metrics are useful with the aims of detecting extreme activity patterns but do not provide the full definitive identification of automated actors. Therefore, the results are not consideredv as fully generalizable through all platforms or communication environments.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec26\" class=\"Section2\"\u003e\n \u003ch2\u003e3.13. Reproducibility and Transparency\u003c/h2\u003e\n \u003cp\u003eTo be able to ensure transparency and reproducibility, all computational experiments were conducted with the utilization of the standardized MATLAB software that also simulates the analytical workflow, which is described in this study.\u003c/p\u003e\n \u003cp\u003eThis software includes the following steps: raw data processing, statistical analysis, behavioral modeling and network simulation.\u003c/p\u003e\n \u003cp\u003eThis computational workflow could provide a transparent methodological framework that make possible for future researchers interested in the topic to replicate the analytical procedures under equivalent conditions for the experiment.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec27\" class=\"Section2\"\u003e\n \u003ch2\u003e3.14. Originality of Figures and Tables\u003c/h2\u003e\n \u003cp\u003eAll figures, tables, graphs, and simulation outputs, which are presented in this study were generated by the author using the computational methods and datasets, that are described in the Methodology section. No figures or tables were reproduced from external publications or electronic sources. All graphs were specifically designed for this study by implementation of the computational analysis based on MATLAB software.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"4. Results","content":"\u003cdiv id=\"Sec29\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Overview of Experimental Design\u003c/h2\u003e \u003cp\u003eA number of computational analyses using MATLAB tool was conducted for examining the characteristics like user engagement, message amplification, highly active accounts and network structures in political communication within the digital world.\u003c/p\u003e \u003cp\u003eThe study is focused on the presidential elections between 2016 and 2024 US, providing a perspective on comparison on the evolution of political discourse online. The datasets include approximately 355,000 tweets from 82,417 users (2016) and 402,000 tweets from 96,210 users (2024).\u003c/p\u003e \u003cp\u003eThe impact scores were calculated through the usage of a composite index based on retweet rates, follower interactions and message propagation within the given network. All current figures and tables are derived from the described analysis on the computational background.\u003c/p\u003e \u003cp\u003eThe results show that the political communication is considered as highly concentrated: a unnoticeable number of highly active accounts tend to generate and amplify a large share of the overall content. Simulation findings further indicate that the automated or highly active actors are more likely to significantly increase the network propagation, which is also highlighting the algorithmic amplification\u0026rsquo;s importance in shaping the digital political discourse.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec30\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Distribution of User Activity\u003c/h2\u003e \u003cp\u003eThe first experiment has examines the distribution of tweets among users, which are participating in election related to discussions. Engagement patterns on social media platforms are generally seen and considered as highly not even, with only given small number of users accounting for the majority of communication activities.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDataset Summary Statistics\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003e Metric\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2016 Election\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2024 Election\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal Tweets\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e355,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e402,000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal Users\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e82,417\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e96,210\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAvg Tweets per User\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.18\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAvg Retweets per Tweet\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8.9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedian Retweets\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh-Activity Users\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4,128\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6,214\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e% High-Activity Users\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.0%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6.5%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe 2016 election data is seen to be contained 355,000 tweets from 82,417 users, which is averaging 4.31 tweets per user. However, user engagement levels are likely to be varied considerably. Approximately 68.4% of users posted fewer than 3 tweets comparing to the percentage of 24.4% posted between 23 and 24.4% posted 20 or more tweets. This suggests that a relatively tiny number participants showing high activity had a disproportionately large impact on the overall political discourse\u0026rsquo; volume.\u003c/p\u003e \u003cp\u003eA similar trend could be observed in the 2024 election data. It contained 402,000 tweets from 96,210 users, averaging 4.18 tweets per user. Approximately 71.6% of users seemed to post fewer than three tweets, at the same time 22.1% posted between 23 and 24.4% posted 20 or more tweets.\u003c/p\u003e \u003cp\u003eThese results revealed a striking long-tailed engagement structure, where a small number of users show a major portion of the political discourse.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis visual representation shows a significantly high imbalance in the engagement\u0026rsquo;s structure, with a small percentage of users who are constituting the majority of communication within the political frame.\u003c/p\u003e \u003cp\u003eOverall, these results tend to provide strong empirical evidence, which are supporting the first hypothesis, which posits that activity related to the political communication on social media follows a highly unequal distribution of engagement.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec31\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Retweet Amplification Dynamics\u003c/h2\u003e \u003cp\u003eThe second experiment shows the behavior of retweeting from the perspective of a mechanism for amplifying messages in the networks of the political communication.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis figure actively shows the relationship between the number of retweets, user engagement and message reaching across three dimensions.\u003c/p\u003e \u003cp\u003eIn the 2016 election data, the average number of retweets was 6.8 (standard deviation 14.2), with a median of 2. The distribution of retweets was highly skewed. Approximately 74.3% of tweets received the number fewer than 5 retweets, while 21.1% received between 5 and 50 retweets. Only 4.6% of tweets received more than 50 retweets, which is indicating that a small number of messages achieved relatively high levels of reach.\u003c/p\u003e \u003cp\u003eDuring the 2024 election period, the dynamics of reach tended to increase. The average number of retweets was 8.9 (standard deviation 18.7), with a median of 3. Approximately 69.8% of tweets received fewer than 5 retweets, 24.5% received between 5 and 50 retweets, and 5.7% received more than 50 retweets.\u003c/p\u003e \u003cp\u003eThese results indicate a high improvement in message dissemination and communication between the two election periods.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec32\" class=\"Section2\"\u003e \u003ch2\u003e4.4. Detection of Bot-Like Behavioral Patterns\u003c/h2\u003e \u003cp\u003eIn the third experiment, author had examined the behavioral characteristics, which are associated with accounts that might probably operate automatically or in coordination with other given accounts. With the aim to identify these patterns, the author calculated a composite activity index, which was combining the frequency of tweets and retweet engagement metrics.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis visual representation indciates that regular users and highly active accounts are seemed to categorized within a behavioral space in the frames of three dimensions, that is defined by tweet frequency and retweet engagement patterns.\u003c/p\u003e \u003cp\u003eAccounts being at above the top 95th percentile of the activity distribution were categorized as exhibiting the patterns of the bot-like behavior.\u003c/p\u003e \u003cp\u003eIn the election data of 2016, approximately 4,128 accounts (5% of users) showed unusually high levels of activity. These accounts posted an average of 36.4 tweets per user, which seem to be significantly higher than the average of approximately 3.1 tweets per user. Furthermore, highly active accounts achieved an average of 21.7 retweets per tweet, comparing it to approximately 5.9 retweets per user.\u003c/p\u003e \u003cp\u003eApplying the same categorization procedure to the 2024 data identified approximately 6,214 accounts (6.5% of users) with similar behavioral characteristics.\u003c/p\u003e \u003cp\u003eThese results are likely to indicate a gradual increase in the number of accounts showing high activity, or accounts that are likely to be automated, between the two election cycles.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec33\" class=\"Section2\"\u003e \u003ch2\u003e4.5. Behavioral Threshold Structure\u003c/h2\u003e \u003cp\u003eIn order to further investigate differences in behaviour within the system of communication, we created a activity space of three-dimensions that combines tweet frequency, retweet participation and composite activity score.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis graphical picture shows the distribution of the activity in behavioural measures and the thresholds used aiming at distinguishing between typical engagement patterns and markedly elevated activity related to behaviour.\u003c/p\u003e \u003cp\u003eTo differentiate the typical user behaviour from the one considered as abnormally high, behavioural thresholds which are corresponding to the 95th percentile of the activity distribution were used.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec34\" class=\"Section2\"\u003e \u003ch2\u003e4.6 Correlation Between Activity and Message Amplification\u003c/h2\u003e \u003cp\u003eAuthor investigated the relationship between user activity and message amplification with the utilization of correlation analysis of key communication metrics.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCorrelation Results\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable Pair\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCorrelation (r)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTweets vs Retweets\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.47\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRetweets vs Influence\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.71\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe analysis in result revealed a relatively moderate to strong positive relationship between communication activities, message amplification and influence.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis graph illustrates the relationships between user activity, engagement levels, and impact measures.\u003c/p\u003e \u003cp\u003eIn order to further analyze these relationships, author estimated a multiple regression model.\u003c/p\u003e \u003cp\u003eInfluence\u0026thinsp;=\u0026thinsp;β0\u0026thinsp;+\u0026thinsp;β1(Activity) + β2(Retweets)\u003c/p\u003e \u003cp\u003eThis model can explain a large part of the dispersion in the user\u0026rsquo;s influence (R\u0026sup2; = 0.58, p \u0026lt; .001), and shows that activity level and retweet amplification are considered strong indicators of influence within the communicational network.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec35\" class=\"Section2\"\u003e \u003ch2\u003e4.7. Bot-Driven Amplification Experiment\u003c/h2\u003e \u003cp\u003eTo assess the probable impact of automated actors on the dissemination of the information under certain large-scaled conditions, a simulation experiment was conducted with the using an expanded artificial communications network with 5,000 users. This large-scale network is seen as an extension of the basic structural model, which is described in Section \u003cspan refid=\"Sec16\" class=\"InternalRef\"\u003e3.5\u003c/span\u003e and enables the analysis of amplification dynamics in communication environments considered more complex.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis graph illustrates the accelerated spread of messages due to the usage of automated accounts.\u003c/p\u003e \u003cp\u003eTwo scenarios were analyzed:\u003c/p\u003e \u003cp\u003eA baseline network which is consisting solely of human users.\u003c/p\u003e \u003cp\u003eA network comprising 400 automated accounts (bots).\u003c/p\u003e \u003cp\u003eThe results showed that human users retweeted an average of 12.7 times per message, while in comparison at the same time automated accounts retweeted an average of 94.2 times per message.\u003c/p\u003e \u003cp\u003eAn independent samples t-test confirmed that this difference was statistically significant.\u003c/p\u003e \u003cp\u003et(5398)\u0026thinsp;=\u0026thinsp;18.64, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001. The use of automated accounts increased the spread of information across the entire network by approximately 38%.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec36\" class=\"Section2\"\u003e \u003ch2\u003e4.8. Network Structural Effects of Bot Amplification\u003c/h2\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e summarizes the structural characteristics of the simulated communication network.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eNetwork metrics\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMetric\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo Bots\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eBots\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage Degree\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7.40\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMax Degree\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e97\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClustering Coefficient\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.050\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.061\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNetwork Density\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.0024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.0031\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAverage Path Length\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.21\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe results indicate that auto-amplification actually improves network connectivity overall with a slight enhancement of local aggregation structures.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e illustrates the configuration of the network after the automated computing system is implemented. The red nodes represent the bot programs, while the blue ones represent the human users.\u003c/p\u003e \u003cp\u003eThe independent samples t-test, when comparing the distributions of node scores could yield a p-value of 0.00257, which is indicating a statistically significant difference between the two network configurations.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e shows that the communication network arrangement follows a thick-tailed distribution. The presence of bots increases the density of nodes with a large number of connections.\u003c/p\u003e \u003cp\u003eThe effect size (Cohen's coefficient d\u0026thinsp;=\u0026thinsp;0.20) indicates that the self-amplification has a significant effect, even in the fact if it is structurally small.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e illustrates the ranking of node centrality in scenarios of the both network, demonstrating that the automated amplification can enhance the importance of already influential nodes.\u003c/p\u003e \u003cp\u003eFigures \u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e depict the structure of the communication network, score distribution and centrality ranking in scenarios related to both networks.\u003c/p\u003e \u003cp\u003eOverall, these results indicate that automated actors improved the ability to disseminate political messages via enhancing the coherence of the network and strengthening the hub structure within the communication system.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec37\" class=\"Section2\"\u003e \u003ch2\u003e4.9. Comparative Analysis of Election Cycles\u003c/h2\u003e \u003cp\u003eComparative analysis revealed that the structural patterns showing several consistencies across two election cycles.\u003c/p\u003e \u003cp\u003eFirst, political engagement remains the highly concentrated realm, with a small number of active users who are producing the majority of content related to election.\u003c/p\u003e \u003cp\u003eSecond, the intensity of message dissemination could increase between the two election cycles, which is indicating a strengthening of the dissemination of information dynamics within the network.\u003c/p\u003e \u003cp\u003eThird, the proportion of active accounts or accounts likely to be automated has gradually increased over the time.\u003c/p\u003e \u003cp\u003eThese findings provide the empirical evidence with the goal to answer research questions four and five, demonstrating that the automated actors can change the structural characteristics regarding the political social networks and enhance their capabilities of information dissemination.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec38\" class=\"Section2\"\u003e \u003ch2\u003e4.10. Influence Density Landscape of Digital Political Communication\u003c/h2\u003e \u003cp\u003eTo incorporate the structural dynamic properties revealed in previous experiments, a three-dimensional model of impact intensity was created.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis graph can clearly illustrate the differences in engagement, information dissemination intensity and automated account activity between the two election cycles of 2016 and 2024 years.\u003c/p\u003e \u003cp\u003eFurthermore, this graph demonstrates the ways how user interaction with message dissemination contributes to the possible creation of concentrated spheres of influence within communication networks.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003e peak intensity refers to the area within the system of the political communications where the influence is most concentrated.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec39\" class=\"Section2\"\u003e \u003ch2\u003e4.11. Galaxy Visualization of Political Communication Networks\u003c/h2\u003e \u003cp\u003eTo illustrate the structural complexity of the environment for the political communication, a three-dimensional representation of the assemblies was created.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e13\u003c/span\u003e the three-dimensional coordinates represent the level of the standard activity, and degrees of participation and influence are taken from the interaction indicators. The resulting cluster structure indicates the potential emergence of distinct communities and centers of influence within the network of the political communication.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Discussion","content":"\u003cp\u003eThis study examined of the digital communication in the political realm via using large-scale data driven form the social media and the analysis of the computational network of the presidential elections related to 2016 and 2024 years. The results show that online discourse in politics is shaped by unequal participation, amplification mechanisms, network centralization and automated activity, which is supporting prior research on the transformations on the structural level in networked communication (Tucker et al., \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Schillemans, \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eEven though the early perspectives emphasized the potential of democratizing in digital platforms, the findings indicate the sharply increasing concentration. A tiny group of highly active users takes the functions of s central nodes, with algorithmic amplification reinforcing their visibility and influence within the network.\u003c/p\u003e \u003cdiv id=\"Sec41\" class=\"Section2\"\u003e \u003ch2\u003e5.1. Concentration of Political Communication\u003c/h2\u003e \u003cp\u003ePolitical communication seems to be highly concentrated among only a small number of users. In 2016, the top 1% produced 18.7% of tweets, increasing to 21.4% in 2024, which is truly indicating growing centralization.\u003c/p\u003e \u003cp\u003eThese patterns align with power-law distributions in digital networks (Barabasi, 2016; Newman, 2018), where influence is concentrated among the so-called central actors. This suggests that theonline political discourse is shaped by the participation inequalities, which in result is leading to emerging power hierarchies rather than fully decentralized communication structures.\u003c/p\u003e \u003c/div\u003e"},{"header":"6. Conclusion","content":"\u003cp\u003eThis study has made a full analysis of political communication through large-scale data and network modeling, focusing on how engagement, amplification and automation can potentially influence the diffusion of information.\u003c/p\u003e \u003cp\u003eAccording to the findings the communication is highly concentrated, amplification is not even and only a small number of messages and users dominate the main visibility. Bot-like accounts, even though they limited in number, contribute disproportionately to the process of content production and spread, while network structures seems to remain strongly centralized.\u003c/p\u003e \u003cp\u003eComparative results are pointing at the intensity on the increasing amplification and the centralization over time. Simulations further demonstrate that even a small numbers of automated actors can significantly enhance the diffusion of the messages.\u003c/p\u003e \u003cp\u003eOverall, political communication within the digital world is driven by interconnected amplification processes involving users, algorithms and automated systems, which is leading to concentrated influence. These results support the framework of algorithmic amplification and artificial influence, which is at the same time highlighting important implications for platform governance, transparency and democratic resilience.\u003c/p\u003e \u003cp\u003eFuture research should extend to cross-platform dynamics, improve the accurate detection of automated actors and further examine the role of AI in shaping the political communication.\u003c/p\u003e"},{"header":"Glossary","content":"\u003cp\u003e\u003cstrong\u003eAlgorithmic Amplification\u003c/strong\u003e - The known process by which the platform algorithms enhancing the visibility of content through the systems \u0026nbsp;of recommendation and mechanisms of the ranking.\u003c/p\u003e\n\u003cp\u003eAlgorithmic Amplification – The process by which algorithms of platform increasing the visibility of content through systems of recommendation and ranking mechanisms.\u003c/p\u003e\n\u003cp\u003eAmplification – The process of increasing the visibility and reach of the content within the digital world through reposting, retweeting, algorithmic promotion \u0026nbsp;or collaborative activities.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAutomated Account\u003c/strong\u003e - A social media account which is managed partially or entirely by scripts \u0026nbsp;based on automation, bots, or algorithmic systems, rather than through the ones with direct human interaction.\u003c/p\u003e\n\u003cp\u003eAutomated Account – Accounts \u0026nbsp;of social media that are managed partially or entirely by automated scripts, bots or algorithmic systems, rather than through the ones with direct human interaction.\u003c/p\u003e\n\u003cp\u003eBot Activity – Observable behavioral patterns that are produced by automated accounts. These include posting frequency, retweeting behavior and collaborative amplification.\u003c/p\u003e\n\u003cp\u003eComputational Communication Model – A systematic framework with the aim of \u0026nbsp;simulating and evaluating communication dynamics by integrating the empirical data in social media with computational analysis.\u003c/p\u003e\n\u003cp\u003eDigital Political Communication – The exchange of information, opinions and political discourse across various platforms in online \u0026nbsp;world and networks of social media.\u003c/p\u003e\n\u003cp\u003eEcho Chamber – A communication environment where individuals are primarily exposed to information that reinforces their existing beliefs.\u003c/p\u003e\n\u003cp\u003eEngagement Metrics – Quantitative metrics used with the aim to measure the interaction \u0026nbsp;of user with digital content (such as likes, retweets, replies, and shares).\u003c/p\u003e\n\u003cp\u003eInfluence Score – A composite metric used with aim to assess the impact of an account within a social network. It is calculated based on engagement frequency, sharing patterns and the reach itself.\u003c/p\u003e\n\u003cp\u003eInformation Diffusion – The process of messages, information and opinions that are spreading across a network through the process of user interactions, such as sharing, reposting and reacting.\u003c/p\u003e\n\u003cp\u003eMisinformation – The information considered as false or misleading that spreads within a digital communication network, not depending on the intent.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNetwork Centrality\u003c/strong\u003e- A structural property of nodes within a network, that is indicating their relative importance or influence within the structure of communication.\u003c/p\u003e\n\u003cp\u003eNetwork Centrality – The structural characteristics of nodes within a network, that are indicating the relative importance or influence of nodes in the structure of communication.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe author was fully responsible for conceptualization, methodology development, data collection, formal analysis, visualization and preparation of the original manuscript as he prepared it by himself.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research received no external funding.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study relies exclusively on publicly available data of social media and does not include any information considered as personally identifiable or personally identifiable, except for publicly available user IDs. This study adheres to the ethical guidelines that are adopted in computational social science and the guidelines for the responsible use of data ofdigital tracking.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eReproducibility Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe computational experiments were performed via the utilization of the analytical software that is based on MATLAB. The systematic workflow and analysis procedures were described in sufficient detail to allow their reproduction via using equivalent datasets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets analyzed in this study are conisdered as publicly available through the Kaggle Research Data Platform, which is a publicly accessible Twitter dataset highly related to elections. This data was originally collected using the Twitter API and is widely used in academic research on communication in digital political.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe author declares no conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors extend their sincere thanks to their colleagues and reviewers for their valuable feedback, which contributed to improving the quality and clarity of this study. They also express their deep gratitude to everyone who contributed data, materials, and other resources necessary for the completion of this research.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAcemoglu D, Restrepo P (2020) Artificial intelligence, automation, and work. J Economic Perspect 34(1):3\u0026ndash;30. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1257/jep.34.1.3\u003c/span\u003e\u003cspan address=\"10.1257/jep.34.1.3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAllcott H, Gentzkow M (2017) Social media and fake news in the 2016 election. J Economic Perspect 31(2):211\u0026ndash;236. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1257/jep.31.2.211\u003c/span\u003e\u003cspan address=\"10.1257/jep.31.2.211\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAllcott H, Gentzkow M, Yu C (2019) Trends in the diffusion of misinformation on social media. Res Politics. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/2053168019848554\u003c/span\u003e\u003cspan address=\"10.1177/2053168019848554\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAltay S, Nielsen RK, Fletcher R (2022) Quantifying the relationship between misinformation exposure and media use. New Media Soc. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/14614448221091216\u003c/span\u003e\u003cspan address=\"10.1177/14614448221091216\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAral S, Eckles D (2019) Protecting elections from social media manipulation. Science 365(6456):858\u0026ndash;861. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.aaw8243\u003c/span\u003e\u003cspan address=\"10.1126/science.aaw8243\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBadawy A, Addawood A, Lerman K, Ferrara E (2019) Characterizing the 2016 Russian IRA influence campaign. Social Netw Anal Min 9., Article 1. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s13278-019-0578-6\u003c/span\u003e\u003cspan address=\"10.1007/s13278-019-0578-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBadawy A, Lerman K, Ferrara E (2018) Who falls for online political manipulation? The Russian Twitter campaign during the 2016 U.S. election. In \u003cem\u003eProceedings of the World Wide Web Conference\u003c/em\u003e (pp. 162\u0026ndash;168). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1145/3308560.3316494\u003c/span\u003e\u003cspan address=\"10.1145/3308560.3316494\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarber\u0026aacute; P (2015) Birds of the same feather tweet together: Bayesian ideal point estimation using Twitter data. Political Anal 23(1):76\u0026ndash;91. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/pan/mpu011\u003c/span\u003e\u003cspan address=\"10.1093/pan/mpu011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarber\u0026aacute; P, Jost JT, Nagler J, Tucker JA, Bonneau R (2015) Tweeting from left to right: Is online political communication more than an echo chamber? Psychol Sci 26(10):1531\u0026ndash;1542. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/0956797615594620\u003c/span\u003e\u003cspan address=\"10.1177/0956797615594620\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeer D (2017) The social power of algorithms. Inform Communication Soc 20(1):1\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/1369118X.2016.1216147\u003c/span\u003e\u003cspan address=\"10.1080/1369118X.2016.1216147\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBender EM, Gebru T, McMillan-Major A, Shmitchell S (2021) On the dangers of stochastic parrots: Can language models be too big? In \u003cem\u003eProceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency\u003c/em\u003e (pp. 610\u0026ndash;623). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1145/3442188.3445922\u003c/span\u003e\u003cspan address=\"10.1145/3442188.3445922\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenedetto A (2025), September 22 Generative AI, disinformation, and electoral integrity: The EU's response to democratic challenges. \u003cem\u003eITSS Verona Magazine, 4\u003c/em\u003e(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.itssverona.it/itss-magazine\u003c/span\u003e\u003cspan address=\"https://www.itssverona.it/itss-magazine\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenkler Y (2019) Networked propaganda and disinformation. \u003cem\u003eHarvard Kennedy School Working Paper\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.14195/2183-5462_34_24\u003c/span\u003e\u003cspan address=\"10.14195/2183-5462_34_24\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenkler Y, Faris R, Roberts H (2018) Network propaganda: Manipulation, disinformation, and radicalization in American politics. Oxford University Press. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/oso/9780190923624.001.0001\u003c/span\u003e\u003cspan address=\"10.1093/oso/9780190923624.001.0001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBennett WL, Segerberg A (2012) The logic of connective action. Inform Communication Soc 15(5):739\u0026ndash;768. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/1369118X.2012.670661\u003c/span\u003e\u003cspan address=\"10.1080/1369118X.2012.670661\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBessi A, Ferrara E (2016) Social bots distort the 2016 U.S. election online discussion. First Monday 21(11). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5210/fm.v21i11.7090\u003c/span\u003e\u003cspan address=\"10.5210/fm.v21i11.7090\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBimber B (2014) Digital media in the Obama campaigns of 2008 and 2012: Adaptation to the personalized political communication environment. J Inform Technol Politics 11(2):130\u0026ndash;150. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/19331681.2014.895691\u003c/span\u003e\u003cspan address=\"10.1080/19331681.2014.895691\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBommasani R, Hudson DA, Adeli E, Altman RB, Arora S, von Arx S, Bernstein MS, Bohg J, Bosselut A, Brunskill E, Brynjolfsson E, Buch S, Card D, Castellon R, Chatterji N, Chen A, Creel K, Davis JQ, Demszky D, Ladhak F (2021) On the opportunities and risks of foundation models. arXiv. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.48550/arXiv.2108.07258\u003c/span\u003e\u003cspan address=\"10.48550/arXiv.2108.07258\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBovet A, Makse HA (2019) Influence of fake news in Twitter during the 2016 election. Nat Commun 10 Article 7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-018-07761-2\u003c/span\u003e\u003cspan address=\"10.1038/s41467-018-07761-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBradshaw S, Howard PN (2019) \u003cem\u003eThe global disinformation order: 2019 global inventory of organised social media manipulation\u003c/em\u003e (Working Paper 2019.2). Oxford Internet Institute, University of Oxford. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://digitalcommons.unl.edu/scholcom/207\u003c/span\u003e\u003cspan address=\"https://digitalcommons.unl.edu/scholcom/207\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrown TB, 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 DM, Wu J, Amodei D (2020) Language models are few-shot learners. arXiv. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.48550/arXiv.2005.14165\u003c/span\u003e\u003cspan address=\"10.48550/arXiv.2005.14165\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBucher T (2018) If\u0026hellip; then: Algorithmic power and politics. Oxford University Press. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/oso/9780190493028.001.0001\u003c/span\u003e\u003cspan address=\"10.1093/oso/9780190493028.001.0001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarlson M (2015) The robotic reporter: Automated journalism. Digit Journalism 3(3):416\u0026ndash;431. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/21670811.2014.976412\u003c/span\u003e\u003cspan address=\"10.1080/21670811.2014.976412\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCinelli M, De Francisci Morales G, Galeazzi A, Quattrociocchi W, Starnini M (2021) The echo chamber effect on social media. \u003cem\u003eProceedings of the National Academy of Sciences, 118\u003c/em\u003e(9), e2023301118. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1073/pnas.2023301118\u003c/span\u003e\u003cspan address=\"10.1073/pnas.2023301118\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCinus F, Minici M, Luceri L, Ferrara E (2024) Exposing cross-platform coordinated inauthentic activity in the run-up to the 2024 U.S. election. arXiv. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.48550/arXiv.2410.22716\u003c/span\u003e\u003cspan address=\"10.48550/arXiv.2410.22716\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConover MD, Ratkiewicz J, Francisco M, Gon\u0026ccedil;alves B, Flammini A, Menczer F (2011) Political polarization on Twitter. In \u003cem\u003eProceedings of the International AAAI Conference on Web and Social Media\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1609/icwsm.v5i1.14126\u003c/span\u003e\u003cspan address=\"10.1609/icwsm.v5i1.14126\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCouldry N, Mejias UA (2019) The costs of connection: How data is colonizing human life and appropriating it for capitalism. Stanford University Press. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1515/9781503609754\u003c/span\u003e\u003cspan address=\"10.1515/9781503609754\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCrawford K, Dobbe R, Fried G, Kaziunas E, Kak A, Mathur V, Richardson R, Schultz J, Schwartz O, West SM, Whittaker M (2018) AI Now 2018 report. AI Now Institute\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCresci S, Di Pietro R, Petrocchi M, Spognardi A, Tesconi M (2017) The paradigm shift of social spambots. In \u003cem\u003eProceedings of the World Wide Web Conference\u003c/em\u003e (pp. 963\u0026ndash;972). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1145/3041021.3055135\u003c/span\u003e\u003cspan address=\"10.1145/3041021.3055135\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDai Z, Yang Z, Yang Y, Carbonell J, Le QV, Salakhutdinov R (2019) Transformer-XL: Attentive language models beyond a fixed-length context. arXiv. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.48550/arXiv.1901.02860\u003c/span\u003e\u003cspan address=\"10.48550/arXiv.1901.02860\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDas D, Clark A (2019) Satire vs fake news: You can tell by the way they say it. IEEE Trans Artif Intell. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/TransAI46475.2019.00012\u003c/span\u003e\u003cspan address=\"10.1109/TransAI46475.2019.00012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDavis CA, Varol O, Ferrara E, Flammini A, Menczer F (2016) BotOrNot: A system to evaluate social bots. In \u003cem\u003eProceedings of the WWW Companion Conference\u003c/em\u003e (pp. 273\u0026ndash;274). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1145/2872518.2889302\u003c/span\u003e\u003cspan address=\"10.1145/2872518.2889302\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEcker UKH, Lewandowsky S, Cook J, Schmid P, Fazio LK, Brashier N, Kendeou P, Vraga EK, Amazeen MA (2022) The psychological drivers of misinformation belief and its resistance to correction. Nat Reviews Psychol 1(1):13\u0026ndash;29. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s44159-021-00006-y\u003c/span\u003e\u003cspan address=\"10.1038/s44159-021-00006-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEnli G (2017) Mediated authenticity in political communication. Inform Communication Soc 20(8):1218\u0026ndash;1233. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/1369118X.2016.1209330\u003c/span\u003e\u003cspan address=\"10.1080/1369118X.2016.1209330\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFerrara E (2017) Disinformation and social bot operations in the run-up to the 2017 French presidential election. Commun ACM 60(7):33\u0026ndash;35. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1145/3097340\u003c/span\u003e\u003cspan address=\"10.1145/3097340\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFerrara E, Varol O, Davis C, Menczer F, Flammini A (2016) The rise of social bots. Commun ACM 59(7):96\u0026ndash;104. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1145/2818717\u003c/span\u003e\u003cspan address=\"10.1145/2818717\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFloridi L, Cowls J, Beltrametti M, Chatila R, Chazerand P, Dignum V, Luetge C, Madelin R, Pagallo U, Rossi F, Schafer B, Valcke P, Vayena E (2018) AI4People\u0026mdash;An ethical framework for a good AI society: Opportunities, risks, principles, and recommendations. Mind Mach 28(4):689\u0026ndash;707. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11023-018-9482-5\u003c/span\u003e\u003cspan address=\"10.1007/s11023-018-9482-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFreelon D, Wells C (2020) Disinformation as political communication. Political Communication 37(2):145\u0026ndash;156. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/10584609.2020.1723755\u003c/span\u003e\u003cspan address=\"10.1080/10584609.2020.1723755\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFuchs C (2017) Social media: A critical introduction, 2nd edn. Sage\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGillespie T (2018) Custodians of the internet: Platforms, content moderation, and the hidden decisions that shape social media. Yale University Press. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.12987/9780300235029\u003c/span\u003e\u003cspan address=\"10.12987/9780300235029\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoodfellow I, Bengio Y, Courville A (2016) \u003cem\u003eDeep learning\u003c/em\u003e. MIT Press. ISBN 0262035618\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGraefe A (2016) \u003cem\u003eGuide to automated journalism\u003c/em\u003e. Tow Center for Digital Journalism. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7916/D80G3XDJ\u003c/span\u003e\u003cspan address=\"10.7916/D80G3XDJ\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrinberg N, Joseph K, Friedland L, Swire-Thompson B, Lazer D (2019) Fake news on Twitter during the 2016 election. Science 363(6425):374\u0026ndash;378. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.aau2706\u003c/span\u003e\u003cspan address=\"10.1126/science.aau2706\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuess AM, Nagler J, Tucker JA (2019) Less than you think: Prevalence and predictors of fake news dissemination on Facebook. Sci Adv 5(1):eaau4586. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/sciadv.aau4586\u003c/span\u003e\u003cspan address=\"10.1126/sciadv.aau4586\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuess A, Nyhan B, Reifler J (2018), January 9 \u003cem\u003eSelective exposure to misinformation: Evidence from the consumption of fake news during the 2016 U.S. presidential campaign\u003c/em\u003e. European Research Council\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHabermas J (2006) Political communication in media society: Does democracy still enjoy an epistemic dimension? Communication Theory 16(4):411\u0026ndash;426. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1468-2885.2006.00280.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1468-2885.2006.00280.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHelberger N, Pierson J, Poell T (2018) Governing online platforms: From contested to cooperative responsibility. Inform Soc 34(1):1\u0026ndash;14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/01972243.2017.1391913\u003c/span\u003e\u003cspan address=\"10.1080/01972243.2017.1391913\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHongladarom S (2023) [Review of the book \u003cem\u003eThe age of surveillance capitalism: The fight for a human future at the new frontier of power\u003c/em\u003e, by S. Zuboff]. \u003cem\u003eAI \u0026amp; Society, 38\u003c/em\u003e(6), 2359\u0026ndash;2361. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00146-020-01100-0\\\u003c/span\u003e\u003cspan address=\"10.1007/s00146-020-01100-0\\\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoward PN (2020) \u003cem\u003eLie machines: How to save democracy from troll armies, deceitful robots, junk news operations, and political operatives\u003c/em\u003e. Yale University Press. ISBN 978-0300250206\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoward PN, Woolley S, Calo R (2018) Algorithms, bots, and political communication in the US 2016 election: The challenge of automated political communication for election law and administration. J Inform Technol Politics 15(2):81\u0026ndash;93. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/19331681.2018.1448735\u003c/span\u003e\u003cspan address=\"10.1080/19331681.2018.1448735\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIyengar S, Hahn KS (2009) Red media, blue media: Evidence of ideological selectivity in media use. J Communication 59(1):19\u0026ndash;39. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1460-2466.2008.01402.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1460-2466.2008.01402.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJungherr A (2016) Analyzing political communication with digital trace data. Springer. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-3-319-20319-5\u003c/span\u003e\u003cspan address=\"10.1007/978-3-319-20319-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJungherr A (2023) Digital campaigning: How digital media change the work of parties and campaign organizations and impact elections. Edw Elgar. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4337/9781789906769.00035\u003c/span\u003e\u003cspan address=\"10.4337/9781789906769.00035\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJungherr A, Rivero G, Gayo-Avello D (2020) Retooling politics: How digital media are shaping democracy. Cambridge University Press. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1017/9781108297820\u003c/span\u003e\u003cspan address=\"10.1017/9781108297820\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJust N, Latzer M (2017) Governance by algorithms: Reality construction by algorithmic selection on the internet. Media Cult Soc 39:238\u0026ndash;258. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/0163443716643157\u003c/span\u003e\u003cspan address=\"10.1177/0163443716643157\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKeller FB, Schoch D, Stier S, Yang J (2020) Political astroturfing on Twitter: How to coordinate a disinformation campaign. Political Communication 37(2):256\u0026ndash;280. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/10584609.2019.1661888\u003c/span\u003e\u003cspan address=\"10.1080/10584609.2019.1661888\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKitchin R (2017) Thinking critically about and researching algorithms. Inform Communication Soc 20(1):14\u0026ndash;29. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/1369118X.2016.1154087\u003c/span\u003e\u003cspan address=\"10.1080/1369118X.2016.1154087\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKreiss D (2016) \u003cem\u003ePrototype politics: Technology-intensive campaigning and the data of democracy\u003c/em\u003e (1st ed.). Oxford University Press. ISBN 978-0199350254\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKreiss D, Lawrence RG, McGregor SC (2018) In their own words: Political practitioner accounts of strategic social media use. Political Communication 35(1):8\u0026ndash;31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/10584609.2017.1334727\u003c/span\u003e\u003cspan address=\"10.1080/10584609.2017.1334727\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKshetri N, Voas J (2017) The economics of fake news. IT Prof 19(6):8\u0026ndash;12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/MITP.2017.4241459\u003c/span\u003e\u003cspan address=\"10.1109/MITP.2017.4241459\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLazer DMJ, Baum MA, Benkler Y, Berinsky AJ, Greenhill KM, Menczer F, Metzger MJ, Nyhan B, Pennycook G, Rothschild D, Schudson M, Sloman SA, Sunstein CR, Thorson EA, Watts DJ, Zittrain JL (2018) The science of fake news. Science 359(6380):1094\u0026ndash;1096. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.aao2998\u003c/span\u003e\u003cspan address=\"10.1126/science.aao2998\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeCun Y, Bengio Y, Hinton G (2015) Deep learning. Nature 521:436\u0026ndash;444. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/nature14539\u003c/span\u003e\u003cspan address=\"10.1038/nature14539\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLoader BD, Mercea D (2011) Networking democracy? Social media innovations and participatory politics. Inform Communication Soc 14(6):757\u0026ndash;769. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/1369118X.2011.592648\u003c/span\u003e\u003cspan address=\"10.1080/1369118X.2011.592648\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMetaxas PT, Mustafaraj E, Gayo-Avello D (2011) How (not) to predict elections using social media. In \u003cem\u003eProceedings of the IEEE International Conference on Privacy, Security, Risk and Trust\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/PASSAT/SocialCom.2011.98\u003c/span\u003e\u003cspan address=\"10.1109/PASSAT/SocialCom.2011.98\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMittelstadt BD, Allo P, Taddeo M, Wachter S (2016) The ethics of algorithms: Mapping the debate. Big Data Soc 3(2):1\u0026ndash;21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/2053951716679679\u003c/span\u003e\u003cspan address=\"10.1177/2053951716679679\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMutz DC (2006) Hearing the other side: Deliberative versus participatory democracy. Cambridge University Press. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1017/CBO9780511617201\u003c/span\u003e\u003cspan address=\"10.1017/CBO9780511617201\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNapoli PM (2014) Automated media: An institutional theory perspective on algorithmic media production and consumption. Communication Theory 24(3):340\u0026ndash;360. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/comt.12039\u003c/span\u003e\u003cspan address=\"10.1111/comt.12039\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNapoli PM (2019) Social media and the public interest. Columbia University. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7312/napo18454\u003c/span\u003e\u003cspan address=\"10.7312/napo18454\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNieborg DB, Poell T, van Dijck J (2025) Platformization. Edward Elgar. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4337/9781035301447.vol3.00051\u003c/span\u003e\u003cspan address=\"10.4337/9781035301447.vol3.00051\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNyhan B (2020) Facts and myths about misinformation. J Economic Perspect 34(3):220\u0026ndash;236. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1257/jep.34.3.220\u003c/span\u003e\u003cspan address=\"10.1257/jep.34.3.220\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNyhan B, Reifler J (2015) The effect of fact-checking on misinformation. Political Communication 32(1):127\u0026ndash;150. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/10584609.2014.969639\u003c/span\u003e\u003cspan address=\"10.1080/10584609.2014.969639\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eO\u0026rsquo;Neil C (2016) \u003cem\u003eWeapons of math destruction: How big data increases inequality and threatens democracy\u003c/em\u003e. Crown Publishers. ISBN 978-0553418811\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOyekunle D, Matthew U, Preston D, Boohene D (2024) Trust beyond technology algorithms: A theoretical exploration of consumer trust and behavior in technological consumption and AI projects. J Comput Commun 12:72\u0026ndash;102. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4236/jcc.2024.126006\u003c/span\u003e\u003cspan address=\"10.4236/jcc.2024.126006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePariser E (2011) \u003cem\u003eThe filter bubble: What the Internet is hiding from you\u003c/em\u003e. Penguin Press. ISBN 978-0-670-92038-9\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePennycook G, Rand DG (2019) Fighting misinformation on social media using crowdsourced judgments of news source quality. \u003cem\u003eProceedings of the National Academy of Sciences, 116\u003c/em\u003e(7), 2521\u0026ndash;2526. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1073/pnas.1806781116\u003c/span\u003e\u003cspan address=\"10.1073/pnas.1806781116\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePorlezza C (2019) Book review: Nicholas Diakopoulos \u003cem\u003eAutomating the news: How algorithms are rewriting the media\u003c/em\u003e. Journalism 20(10):1417\u0026ndash;1419. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/1464884919876179\u003c/span\u003e\u003cspan address=\"10.1177/1464884919876179\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrior M (2007) \u003cem\u003ePost-broadcast democracy: How media choice increases inequality in political involvement and polarizes elections\u003c/em\u003e. Cambridge University Press. ISBN 978-0-521-67533-8\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePrior M (2007) Post-broadcast democracy: How media choice increases inequality in political involvement. Cambridge University Press\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRamanaharan R, Guruge DB, Agbinya J (2025) Deepfake video detection: Insights into model generalisation\u0026mdash;A systematic review. Data Inform Manage 9 Article 100099. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.dim.2025.100099\u003c/span\u003e\u003cspan address=\"10.1016/j.dim.2025.100099\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoozenbeek J, van der Linden S (2019) Fake news game confers psychological resistance against online misinformation. Palgrave Commun 5., Article 65. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1057/s41599-019-0279-9\u003c/span\u003e\u003cspan address=\"10.1057/s41599-019-0279-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRussell S, Norvig P (2021) \u003cem\u003eArtificial intelligence: A modern approach\u003c/em\u003e (4th ed.). Pearson. ISBN 978-0-13-461099-3\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSayers JG (2014) Book review: Communication power by M. Castells. Work Employ Soc 28(1):142\u0026ndash;144. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/0950017013511515\u003c/span\u003e\u003cspan address=\"10.1177/0950017013511515\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchillemans T (2014) The hybrid media system: Politics and power. Public Adm 92(4):867\u0026ndash;883. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/padm.12108\u003c/span\u003e\u003cspan address=\"10.1111/padm.12108\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShao C, Ciampaglia GL, Varol O, Yang K-C, Flammini A, Menczer F (2018) The spread of low-credibility content by social bots. Nat Commun 9:4787. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41467-018-06930-7\u003c/span\u003e\u003cspan address=\"10.1038/s41467-018-06930-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith D, Ehrett C, Warren P (2025) Unsupervised detection of coordinated information operations in the wild. EPJ Data Sci 14(1):26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1140/epjds/s13688-025-00544-y\u003c/span\u003e\u003cspan address=\"10.1140/epjds/s13688-025-00544-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStarbird K (2017) Examining the alternative media ecosystem through the production of alternative narratives of mass shooting events on Twitter. In \u003cem\u003eProceedings of the International AAAI Conference on Web and Social Media\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1609/icwsm.v11i1.14878\u003c/span\u003e\u003cspan address=\"10.1609/icwsm.v11i1.14878\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStella M, Ferrara E, De Domenico M (2018) Bots increase exposure to negative and inflammatory political content in online social systems. \u003cem\u003eProceedings of the National Academy of Sciences, 115\u003c/em\u003e(49), 12435\u0026ndash;12440. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1073/pnas.1803470115\u003c/span\u003e\u003cspan address=\"10.1073/pnas.1803470115\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStroud NJ (2011) Niche news: The politics of news choice. Oxford University Press. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/acprof:oso/9780199755509.001.0001\u003c/span\u003e\u003cspan address=\"10.1093/acprof:oso/9780199755509.001.0001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSubrahmanian VS, Azaria A, Durst S, Kagan V, Galstyan A, Lerman K (2016) The DARPA Twitter bot challenge. Computer 49(6):38\u0026ndash;46. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/MC.2016.183\u003c/span\u003e\u003cspan address=\"10.1109/MC.2016.183\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSunstein CR (2017) #Republic: Divided democracy in the age of social media. Princeton University Press. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2307/j.ctv8xnhtd\u003c/span\u003e\u003cspan address=\"10.2307/j.ctv8xnhtd\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTheocharis Y, Lowe W, van Deth JW, Garc\u0026iacute;a-Albacete GM (2015) Using Twitter to mobilize protest action. Inform Communication Soc 18(2):202\u0026ndash;220. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/1369118X.2014.948035\u003c/span\u003e\u003cspan address=\"10.1080/1369118X.2014.948035\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTucker JA, Guess A, Barber\u0026aacute; P, Vaccari C, Siegel A, Sanovich S, Stukal D, Nyhan B (2018) Social media, political polarization, and political disinformation: A review of the scientific literature. SSRN Electron J. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2139/ssrn.3144139\u003c/span\u003e\u003cspan address=\"10.2139/ssrn.3144139\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVaccari C, Chadwick A (2020) Deepfakes and disinformation: Exploring the impact of synthetic political video on deception, uncertainty, and trust in news. Social Media + Soc. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/2056305120903408\u003c/span\u003e\u003cspan address=\"10.1177/2056305120903408\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan Dijck J, Poell T, de Waal M (2018) The platform society: Public values in a connective world. Markets Globalization Dev Rev 3(3). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eArticle 6. https://doi.org/10.23860/MGDR-2018-03-03-08\u003c/span\u003e\u003cspan address=\"Article 6. 10.23860/MGDR-2018-03-03-08\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVosoughi S, Roy D, Aral S (2018) The spread of true and false news online. Science 359(6380):1146\u0026ndash;1151. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.aap9559\u003c/span\u003e\u003cspan address=\"10.1126/science.aap9559\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWardle C, Derakhshan H (2017) \u003cem\u003eInformation disorder: Toward an interdisciplinary framework for research and policy making\u003c/em\u003e (DGI(2017)09). Council of Europe. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://rm.coe.int/information-disorder-toward-an-interdisciplinary-framework-for-researc/168076277c\u003c/span\u003e\u003cspan address=\"https://rm.coe.int/information-disorder-toward-an-interdisciplinary-framework-for-researc/168076277c\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWoolley SC (2020) \u003cem\u003eThe reality game: How the next wave of technology will break the truth, and what we can do about it\u003c/em\u003e. Endeavour. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.worldcat.org/oclc/1127085781\u003c/span\u003e\u003cspan address=\"https://www.worldcat.org/oclc/1127085781\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWoolley SC, Howard PN (2017) \u003cem\u003eComputational propaganda worldwide: Executive summary\u003c/em\u003e (Working Paper No. 2017.11). University of Oxford\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZannettou S, Sirivianos M, Blackburn J, Kourtellis N (2019) The web of false information. In \u003cem\u003eProceedings of the ACM Internet Measurement Conference\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1145/3309699\u003c/span\u003e\u003cspan address=\"10.1145/3309699\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou X, Zafarani R (2021) A survey of fake news: Fundamental theories, detection methods, and opportunities. ACM-CSUR 53(5):1\u0026ndash;40. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1145/3395046\u003c/span\u003e\u003cspan address=\"10.1145/3395046\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"humanities-and-social-sciences-communications","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"palcomms","sideBox":"Learn more about [Humanities \u0026 Social Sciences Communications](http://www.nature.com/palcomms/)","snPcode":"41599","submissionUrl":"https://submission.springernature.com/new-submission/41599/3","title":"Humanities and Social Sciences Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Digital political communication, computational social science, information diffusion, algorithmic amplification, automated accounts, network centralization","lastPublishedDoi":"10.21203/rs.3.rs-9473319/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9473319/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn this study computational social science methods are implemented to large-scale election datasets. The present study aims to examine the structural dynamics of the political communication within the landscape of social media. At the same time it conducts a comparative analysis of patterns related to communication in the 2016 and 2024 US presidential elections, which focus on participation disparities, message amplification, automated behavior, and network architecture.\u003c/p\u003e \u003cp\u003eAlongside with the study\u0026rsquo;s framework, statistical analysis, network modeling and MATLAB simulations are used aiming at examining the ways of how political information spreads and gains visibility in social media realm. The results can be interpreted as the ones with clear imbalances. It should be noted, that only a tiny group of users produces most of the content seen in social media, while others tend to remain less active. Message usage can be seen as not even and only a few messages could in result achieve wide dissemination.\u003c/p\u003e \u003cp\u003eThe findings also indicate that bot-like behavioral characteristics, which indicate that these bots are inclined to produce and disseminate content at a much higher rate comapring to the average user. A notable pattern emerges indicating that actors with strong connections act in the roles of central hubs and highly contribute to shaping the dissemination of the information. The comparative analysis showed a significant relationship between that dissemination power, network centralization, and the reach of active accounts, which increased in 2024.\u003c/p\u003e \u003cp\u003eIt is worth noting that the results of the current simulation indicated that even with a small number of automated accounts, they can significantly increase information reach.\u003c/p\u003e","manuscriptTitle":"Truth, Trust, and Technology: Political Communication in the Era of Synthetic Media","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-11 05:10:59","doi":"10.21203/rs.3.rs-9473319/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-05-07T11:42:47+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-05-06T23:41:38+00:00","index":"","fulltext":""},{"type":"submitted","content":"Humanities and Social Sciences Communications","date":"2026-05-06T17:01:22+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"humanities-and-social-sciences-communications","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"palcomms","sideBox":"Learn more about [Humanities \u0026 Social Sciences Communications](http://www.nature.com/palcomms/)","snPcode":"41599","submissionUrl":"https://submission.springernature.com/new-submission/41599/3","title":"Humanities and Social Sciences Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"620fabef-2d4f-4534-989b-0f7ea693be2c","owner":[],"postedDate":"May 11th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Revision requested","date":"2026-05-07T11:42:47+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-05-06T23:41:38+00:00","index":"","fulltext":""},{"type":"submitted","content":"Humanities and Social Sciences Communications","date":"2026-05-06T17:01:22+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[{"id":67710635,"name":"Humanities/Complex networks"},{"id":67710636,"name":"Social science/Complex networks"},{"id":67710637,"name":"Physical sciences/Mathematics and computing"},{"id":67710638,"name":"Physical sciences/Physics"}],"tags":[],"updatedAt":"2026-05-11T05:10:59+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-11 05:10:59","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9473319","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9473319","identity":"rs-9473319","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}