What types of COVID-19 conspiracies are populated by Twitter bots?

“tc.). The presence of bots has been documented in the context of online political discussion [1]–[3], public health [4]–[8], civil unrest [9], stock market manipulation [10], the spread of false news [11]–[13], alongside with other tools such as troll accounts [5], [6], [14]–[17]. In this paper, we chart the landscape…”

“tion on February 11, 2020) were reported in Wuhan, China in late December 2019; the first fatalities were reported in early 2020. The first case in the United States was announced on January 21, 2020 [18]: our Twitter data collection aligns with that date [19]. The fast-rising infection rates and death toll led the Ch…”

“On a different note, [23] looked at social media sentiment as expression of potential mental health issues associated with social isolation and other side-effects of containment measures enacted to limit the spread of COVID-19 in China.”

“ve a digital trail associated with the data center used to dispatch tweets. RQ2: CHARACTERIZING BOT FUELED CAMPAIGNS Data-driven computational sensemaking can fail when using human communication data [66]. Common reasons include population and sampling bias, platform-design bias, distortion of human and nonhuman behav…”

“[74] M.”

“The literature on bot detection has become very extensive [31], [36], [49]–[51].”

“[12] N.”

“The work by Pennycook and colleagues epitomizes the seriousness of this problem by showing, with a social experiment including 1,600 participants, that subjects tend to share misinformation and false claims about COVID-19 predominantly because they are unable to determine whether the content is scientifically sound an…”

“According to work by [5], [42], similar conclusions can be drawn for bots active in public health discussions, with the intent to spread claims contrary to scientific evidence.”

“ed by the activity of automated accounts, often referred to as bots (a.k.a., social bots, sybil accounts, etc.). The presence of bots has been documented in the context of online political discussion [1]–[3], public health [4]–[8], civil unrest [9], stock market manipulation [10], the spread of false news [11]–[13], a…”

“(1) Botometer3, also used here, is a bot-detection tool developed at Indiana University [52]; (2) BotSlayer4 is an application to detect and track potential manipulation of information on Twitter; (3) the Bot Repository5 is a centralized database to share annotated datasets of Twitter bots.”

“idation section. The spread of conspiracies on online social media is a well-established issue. Numerous studies have been devoted to understand, for example, how unscientific claims circulate online [71]–[74], or how conspiratorial narratives are constructed online [75]–[77], especially in the context of political id…”

“nly observing a small fraction, approximately 1%, of the overall Twitter conversation, through the lens of the Twitter API. This has been shown to introduce some biases toward over-represented topics [87], and COVID-19 has been the most spoken-about topic of discussion ever since the beginning of the pandemic. Second,…”

“Numerous studies have been devoted to understand, for example, how unscientific claims circulate online [71]–[74], or how conspiratorial narratives are constructed online [75]–[77], especially in the context of political ideology [2], [78].”

“ation and sampling bias, platform-design bias, distortion of human and nonhuman behavior, multiple hypotheses and comparisons testing, and more, which can hamper computational models and bias results [67]. An hybrid approach based on combining computational analysis and human validation has been called for in order to…”

“[72] A.”

“oted to understand, for example, how unscientific claims circulate online [71]–[74], or how conspiratorial narratives are constructed online [75]–[77], especially in the context of political ideology [2], [78]. Our analysis highlights the emergence of similar issues in the context of the conversation related to COVID-…”

“centralized database to share annotated datasets of Twitter bots. Finally, various models have been proposed to detect bots using sophisticated machine learning techniques, such as: (1) Deep learning [53], (2) Anomaly detection [54]–[57], (3) Time series analysis [38], [58], [59]. Due to the continuously evolving natu…”

“en referred to as bots (a.k.a., social bots, sybil accounts, etc.). The presence of bots has been documented in the context of online political discussion [1]–[3], public health [4]–[8], civil unrest [9], stock market manipulation [10], the spread of false news [11]–[13], alongside with other tools such as troll accou…”

“[4] V.”

“nced Research Projects Agency) aimed at detecting bots used for anti-science misinformation [4]. More recently, large bot networks (botnets) have been discovered on Twitter by various research groups [43], [44], [46], [48]. The literature on bot detection has become very extensive [31], [36], [49]–[51]. In [31], we pr…”

“esearch Projects Agency) aimed at detecting bots used for anti-science misinformation [4]. More recently, large bot networks (botnets) have been discovered on Twitter by various research groups [43], [44], [46], [48]. The literature on bot detection has become very extensive [31], [36], [49]–[51]. In [31], we proposed…”

“e identification of highly-predictive features that discriminate between bots and humans. Other recent surveys propose complementary or alternative taxonomies that are worth considering as well [36], [50], [51] Some openly accessible tools exist to detect bots on platforms like Twitter: (1) Botometer3, also used here,…”

“An hybrid approach based on combining computational analysis and human validation has been called for in order to make sense of massive-scale communication datasets [68], [69].”

“inuum of behavior along the dimensions of social media engagement is also in line with two other theories, namely the notions of “super-participation” [61] and that of “dynamical classes of behavior” [62]. In our case, we adopt the same idea to segmenting the dimension of human-to-bot likeness and dividing the continu…”

“[15] A.”

“[73] D.”

“r further details. Bot behavior research. Numerous studies have been devoted to characterize the behavior of bots [36]– [38] and the ethical issues pertaining their use [39]. E.g.: According to [30], [40], [41], the behavior of political bots can be categorized with respect to their goals: • Manufacturing consensus, t…”

“is a well-established issue. Numerous studies have been devoted to understand, for example, how unscientific claims circulate online [71]–[74], or how conspiratorial narratives are constructed online [75]–[77], especially in the context of political ideology [2], [78]. Our analysis highlights the emergence of similar…”

“35] for further details. Bot behavior research. Numerous studies have been devoted to characterize the behavior of bots [36]– [38] and the ethical issues pertaining their use [39]. E.g.: According to [30], [40], [41], the behavior of political bots can be categorized with respect to their goals: • Manufacturing consen…”

“Bots have been categorized according to various taxonomies [29], [30].”

“ch social media. The work by Singh and colleagues also looked at the spatio-temporal dynamics of misinformation spread on Twitter, drawing a picture with similar implications as the two studies above [26]. More research is needed, as the information landscape evolves, and more scientific insights are unveiled on the c…”

“ng bots appears to dissipate toward the end of our observation window (early March). This type of bot behavior has been documented already in the past, and is commonly related to as “trend hijacking” [85], [86]: bots appear to ride the wave of popularity of a given topic, in this case the trending COVID-19 discussion,…”

“odels and bias results [67]. An hybrid approach based on combining computational analysis and human validation has been called for in order to make sense of massive-scale communication datasets [68], [69]. Content sensemaking. In this section, we address RQ2 using content and timeseries analysis techniques, in combina…”

“[45] Y.”

“The presence of bots has been documented in the context of online political discussion [1]–[3], public health [4]–[8], civil unrest [9], stock market manipulation [10], the spread of false news [11]–[13], alongside with other tools such as troll accounts [5], [6], [14]–[17].”

“As of the time of this writing (mid-April 2020), the vast majority of these studies are pre-print papers that provide a timely, yet partial, characterization of online discussion and issues revolving around COVID-19 [19]–[28].”

“The idea that there exists a continuum of behavior along the dimensions of social media engagement is also in line with two other theories, namely the notions of “super-participation” [61] and that of “dynamical classes of behavior” [62].”

“h Projects Agency) aimed at detecting bots used for anti-science misinformation [4]. More recently, large bot networks (botnets) have been discovered on Twitter by various research groups [43], [44], [46], [48]. The literature on bot detection has become very extensive [31], [36], [49]–[51]. In [31], we proposed a sim…”

“tweet text alongside with the URL. The large majority of the observed news bots tweets have been captured because the headlines contain one of the keywords used to seed the real-time data collection [19]. DISCUSSION COVID-19 is a global crisis and with people being pushed out of physical spaces due to containment meas…”

“The presence of bots has been documented in the context of online political discussion [1]–[3], public health [4]–[8], civil unrest [9], stock market manipulation [10], the spread of false news [11]–[13], alongside with other tools such as troll accounts [5], [6], [14]–[17].”

“[59] D.”

“[7] T.”

“[37] L.”

“[16] A.”

“[56] Z.”

“[34] E.”

“well-established issue. Numerous studies have been devoted to understand, for example, how unscientific claims circulate online [71]–[74], or how conspiratorial narratives are constructed online [75]–[77], especially in the context of political ideology [2], [78]. Our analysis highlights the emergence of similar issue…”

“[88] E.”

“[76] A.”

“y an alleged "deep state" against U.S. President Donald Trump and his supporters”. QAnon has been extensively adopted by alt-right activists to foster participatory advocacy on social media [79], but it has also been abused by the Russian Internet Research Agency (IRA) to push conspiratorial and divisive nar…”

“[84] D.”

“The presence of bots has been documented in the context of online political discussion [1]–[3], public health [4]–[8], civil unrest [9], stock market manipulation [10], the spread of false news [11]–[13], alongside with other tools such as troll accounts [5], [6], [14]–[17].”

“to determine whether the content is scientifically sound and accurate or not [22]. Other studies investigated collective attention and engagement dynamics concerning COVID-19 on Twitter. For example, [20] analyzed 1,000 unigrams (1-grams) posted on Twitter in 24 languages during early 2020, and compared them with the…”

“(1) Deep learning [53], (2) Anomaly detection [54]–[57], (3) Time series analysis [38], [58], [59].”

“e annotated datasets of Twitter bots. Finally, various models have been proposed to detect bots using sophisticated machine learning techniques, such as: (1) Deep learning [53], (2) Anomaly detection [54]–[57], (3) Time series analysis [38], [58], [59]. Due to the continuously evolving nature of bots, and the challeng…”

“s appears to dissipate toward the end of our observation window (early March). This type of bot behavior has been documented already in the past, and is commonly related to as “trend hijacking” [85], [86]: bots appear to ride the wave of popularity of a given topic, in this case the trending COVID-19 discussion, to in…”

“ects Agency) aimed at detecting bots used for anti-science misinformation [4]. More recently, large bot networks (botnets) have been discovered on Twitter by various research groups [43], [44], [46], [48]. The literature on bot detection has become very extensive [31], [36], [49]–[51]. In [31], we proposed a simple ta…”

“The presence of bots has been documented in the context of online political discussion [1]–[3], public health [4]–[8], civil unrest [9], stock market manipulation [10], the spread of false news [11]–[13], alongside with other tools such as troll accounts [5], [6], [14]–[17].”

“Numerous studies have been devoted to characterize the behavior of bots [36]– [38] and the ethical issues pertaining their use [39].”

“[80] G.”

“Numerous studies have been devoted to characterize the behavior of bots [36]– [38] and the ethical issues pertaining their use [39].”

“, GreatAwakening, Inf0wars, WWGWGA (Where We Go One We Go All), WeThePeople, and PatriotsFights [82].”

“[70] F.”

“nline political discussion [1]–[3], public health [4]–[8], civil unrest [9], stock market manipulation [10], the spread of false news [11]–[13], alongside with other tools such as troll accounts [5], [6], [14]–[17]. In this paper, we chart the landscape of Twitter chatter within the context of COVID-19 related convers…”

“algorithmic manipulation, to trick recommendation and ranking systems used by social media platforms, and give higher visibility to certain actors, viewpoints, or campaigns. According to work by [5], [42], similar conclusions can be drawn for bots active in public health discussions, with the intent to spread claims c…”

“Similar approaches have been developed in the domain of information warfare detection: one such example, is the CUT (Commenting User Typology) Framework, which was proposed to classify comment abuse, where users are divided into quadrants of behaviors alongside various dimensions of interest [60].”

“t-right activists to foster participatory advocacy on social media [79], but it has also been abused by the Russian Internet Research Agency (IRA) to push conspiratorial and divisive narratives [80], [81]. QAnon appears alongside with other known alt-right terms, e.g., GreatAwakening, Inf0wars, WWGWGA (Where We Go One…”

“nt, or clonedcontent from other users; finally, (7) pausing activity to mimic human circadian cycles and bursty behaviors, as well as to satisfy API constraints, and to avoid suspension. According to [35], these features can enable bots to carry out various forms of abuse including: denigrate and smear, deceive and ma…”

“In an analysis of 43 million COVID-19–related tweets, bots were found to be pushing a number of conspiracy theories, such as QAnon, in addition to retweeting links from partisan news sites [73].”

“Bots also actively participate in public health debates [12] including those about vaccines [13, 14], the COVID-19 pandemic [15, 16, 17, 18], and cannabis [19].”

“Therefore, similar to previous work [23,26,28], we binned users by evenly distributed deciles of the polarity scores, with each decile containing exactly 10% of all users.”

“…social bots can pose significant security risks by influencing public opinion and disseminating false information (Shao et al. 2017), spreading rumors and conspiracy theories (Ferrara 2020), creating fake reputations, and suppressing political competitors (Pierri et al. 2020; Benkler et al. 2017).”

“Recent work using our dataset has already begun to explore the prevalence of bots and misinformation within the 2020 political landscape [6, 7].”

“Bots are automated accounts that post only certain types of information, often in collaboration with other bots, to boost the visibility of certain hashtags, news sources, or storylines (Ferrara, 2020).”

“The COVID-19 pandemic presents a novel chance to assess where health misinformation becomes political (Ferrara, 2020).”

“…demonstrated that accounts with a high “bot score” (indexing the likelihood of being a bot, or fake account) promulgate conspiratorial narratives charged with alt-right ideology and are specifically oriented toward hateful and polarizing political ideologies (Stella et al., 2018; Ferrara, 2020).”

“Ferrara (2020) studies the role of bots in amplifying COVID-19 conspiracy theories.”

“Analysis has also revealed that “high bot score accounts are used to promote political conspiracies and divisive hashtags alongside with COVID-19 content” (Ferrara 2020 p. 17), while accounts likely run by humans focus more on health and public welfare.”

“There is ample scientific knowledge of bot activity in online talk about the COVID-19 pandemic, especially regarding their spread of COVID-19 conspiracy theories (Ferrara, 2020; Moffitt et al., 2021) and more general low-credibility information (Himelein-Wachowiak et al.”

“Ferrara [20], for instance, used bot detection techniques to study 43.3 million COVID-19 related English tweets and observed how bots used COVID-19 as a vector to promote ideological hashtags that were often associated with the alt-right in the USA.”

“Furthermore, during the COVID-19 pandemic, there has been an uptick in rumours and conspiracies spreading through online platforms(Ferrara, 2020).”

“Ferrara [10] analyzed 43.3 million English tweets and found the preferences of bots and humans in tweet contents by analyzing n-grams and hashtags.”

“, social bots [68], and an investigation to relate bot behavior to suspension is left for future research, considering that a correlation between account suspension and increased bot likeness was already observed in other social and political discussions [69, 70].”

“Also, artificially created information by bots has an impact (Ferrara, 2020).”

“Bots have reportedly influenced the 2016 US Presidential Election [5, 38], the Brexit vote in the UK [4, 38], the spread of misinformation about COVID-19 [27] and financial markets [13, 54].”

“The active high-cited paper, by Ferrara [18] in 2020, demonstrates how social media tools such as Twitter were used to prompt political conspiracies amid the C-19 catastrophe.”

“In line with recent literature (Ferrara, 2020; Yang et al., 2019), we used the top 10 and bottom 10 percentiles of users to set apart likely human users from automated accounts, with Botscores of 0.06 and below for humans and 0.67 and above for bots.”

“In this context, hyper-partisan and alternative media platforms may contest conventional knowledge of mainstream media and politicians (e.g., Heft et al., 2020), or even disseminate conspiracies emphasizing alternative interpretations (Ferrara, 2020).”

“…other coordinated networks, again, are theorized to be employed for different purposes, ranging from the spread of CTs and political disinformation (Ferrara, 2020; Giglietto et al., 2020; Graham et al., 2020; Shao et al., 2017) or simulating mass support for a political matter (Elmas et al.,…”

“Bots – defined here as co-ordinated internet agents that simulate human activity to share content – have also played an important role, with greater likelihood of sharing misinformation (Ferrara, 2020; Yang et al., 2020).”

“In the context of COVID-19, there was an extensive application of bots to influence public opinion by fueling conspiracy theories and promoting divisive hashtags [48], [49].”

“The dataset was collected by using Twitter API dated from 22 January 2020 to 13 March 2020.”

“Researchers observed how Twitter bot accounts used #COVID-19 to promote ideological hashtags typically associated with the U.S. ’ s alt-right values (Ferrara, 2020).”

“A wealth of research has been conducted in the past two years to better understand the dynamics of COVID-19 related misinformation and its effect on our society and on public health measures [18, 26, 33, 51, 75, 76].”

“…pandemic with a myriad of methodologies and objectives: among others, Mackey et al. (2020) use Twitter data to research symptoms associated with the disease, Pulido et al. (2020) pay attention to the spread of false information, and Ferrara (2020) focuses on the role played by bots in such process.”

“Researchers have raised concerns about misinfor-mation, myths, and conspiracies related to COVID-19 [4]–[6], [14], [15].”

“Further efforts aimed at quantifying the magnitude and the effects of infodemic on social media investigating a broader amount of platforms [5, 27, 28].”

“This tool has been widely applied to detect social bots tweeting about coronavirus misinformation (Ahmed et al., 2020; Ferrara, 2020; Gruzd and Mai, 2020; Jamison et al., 2020).”

“The bot detection literature is vast (Ferrara 2020), and a bot is usually identified based on their overall posting behavior during a specified period of time.”

“Bots have often been implicated in the spread of misinformation online (Bastos & Mercea, 2018, 2019; Broniatowski et al., 2018; Ferrara, 2020), and are often driven by state actors (Alsmadi & O’Brien, 2020; Prier, 2017).”

“…anti-vaccine propaganda was mostly supported with false narratives and conspiracies, and the prevalence of anti-vaccine misinformation was conspicuous on social media platforms (Cha et al., 2021; Ferrara, 2020a; Gallotti et al., 2020; Gruzd et al., 2021; S. K. Lee et al., 2022; Singh et al., 2022).”

“Previous works have focused on analyzing online behavior and reactions to the spread of COVID-19 [6], or investigating conspiracy theories and social activism [8], [9].”

“Examples include the spread of misinformation on Twitter [12,13], emotional consequences of lockdowns studied through Twitter (and Weibo) data [14], as well as the association between US partisanship and sentiment towards COVID-19 measures [15].”

“Naturally, this amplified the discussions around Covid-19 vaccines both offline and online.”

“They spread misinformation about COVID-19 [10], [11] and financial markets [12], [13].”

“Throughout the COVID-19 pandemic [16], bots spread misinformation about the virus and the vaccines on social media, leading to mob panic, confusion, and even resistance to public health measures.”

“3, 4, 5, 6, 7 and 8 for various metrics.”

“Similarly, in [18] and [19], the authors analyzed the spread of coronavirus related misinformation due to bots and other automated accounts.”

“…et al., 2021; Weber et al., 2022; Yang et al., 2021; Zola et al., 2022) some studies found evidence of suspected coordinated, inauthentic behaviour (Ferrara, 2020; Graham et al., 2020, 2021; Graham & Keller, 2020; Keller et al., 2020), with far-right accounts (QAnon and MAGA accounts) being identi…”

“In contrast to other contexts, such as the study by Duan et al. (2022) and Ferrara (2020), Iranian users (bot or human) were less inclined to propagate conspir-acy theories during this politically contentious period, choosing instead to channel their efforts into their political goals.”

“Twitter received masses of tweets from all across the world about Covid-19 and its related problems [2].”

“The malicious usage of bots was the source of misinformation on social media [1], spread of conspiracy theories [2], and political election interference [3].”

“…fast in addressing several issues connected to misinformation, such as its impact on opinion formation (Shahi et al., 2021), conspiracy theories, and hate speech (Tahmasbi et al., 2021), the power of automated bots and accounts (Ferrara, 2020), and the impact on human rights (Abrusci et al., 2020).”

“(Ferrara, 2020; see Rodríguez-Ruiz et al., 2020; Shevtsov et al., 2020).”

“Several studies show that bots and ordinary users promote misinformation on social media [6,9,18,41].”

“However, in [35]–[37], the authors argue that a substantial percentage of conspiratorial information remains online on YouTube and other platforms, influencing the public despite so much research.”

“The 2016 United States Presidential Election [6], [13], [14], The 2018 United States Midterm Election [15], [16], [17], Germany’s Federal Election [18], [7], French Presidential Election [8] and large political events around the world [20], [21], [27].”

“There are a number of topics fueling COVID-related misinformation, ranging from conspiracy theories, misreporting of morbidity and mortality, disease spread mechanisms, prevention meth-ods, treatments and drugs, recovery experiences, and political controversies [3], [4].”