Crowdfunding (as) disinformation: ‘Pitching’ 5G and election fraud campaigns on GoFundMe

While funding for online disinformation research has dramatically increased after the election of Donald Trump, the financial support and fundraising for disinformation is rarely addressed in the most high-profile research reports. The Citizen Lab’s comprehensive ‘Disinformation Annotated Bibliography’ (Lim, 2019) for example includes remarkably few discussions of disinformation campaign finances, and zero references to crowdfunding. The bibliography breaks up the report into nine distinct themes, none of which address the financial support for disinformation. Marwick and Lewis (2017) and Allcott and Gentzkow’s (2017) though do note potential economic motivations. In both cases however the remarks are restricted to discussions of the attention economy of social media (Wu, 2017) and how incorrect information can be leveraged through advertising networks to create income for individuals. Marwick and Lewis also note the demise of newspaper funding and budgets as another factor in the spread of disinformation. Bakir and McStay’s (2018) offer a more fulsome review of this perspective, arguing that the economic changes in the news industry coupled with social media algorithms have created a space for the production and circulation of disinformation. And to this ecology, Misiewicz and Yu (2017) add the growing importance of misleading advertising and fraud that has arisen through journalism related algorithms.

Throughout the literature there are analyses of many internet platforms, yet not a single reference to crowdfunding sites – even previews of future trends fail to address crowdfunding, choosing instead to focus on texting applications and virtual reality interfaces. (Wardle and Derakhshan, 2018) That said, two reports included in the bibliography do offer important insights into crowdfunding as disinformation. While not focused on crowdfunding per se, Gu et al. (2017) investigates how the VTope crowdsourcing platform rewards users who circulate and otherwise spread false information across the internet. Participants on the Russian platform – that boasts over two million users – collect points that can be cashed in or used for reputational gain.

Nadler et al. (2018) also highlight the toxic influence of ‘dark money’ in contributing to a crisis of democratic disinformation. The authors discuss how such resources are used to weaponize messages, though like other reports in the bibliography the authors limit their analysis to the advertising industry. The relevance of Nadler et al.’s study for our study lies in its discussion of the ‘weaponization’ of political speech and language, again particularly as it targets and blames particular demographics (immigrants, racialized communities) and political opponents online. (Ott and Dickenson, 2019) Given that ‘pitches’ are such a central component of crowdfunding campaigns – both visually on the platform and in framing the terms of the fundraising appeal – their communicative role in spreading disinformation is a rather glaring blind spot in the literature to date. Due to the absence of such concerns for communication and fundraising in general in the disinformation research we turn next to the fundraising literature from political communication to frame our study of crowdfunding – and in particular their ‘networked pitches’ – as potential forms of disinformation.

A common thread in studies of political fundraising is the shift to increasingly personalized fundraising appeals and technologies. The personalization turn in fundraising occurred in the late 1970s with the development of the ‘Total Design Method’, an approach that sought to increase fundraising engagement by personalizing names of recipients, using real signatures, and sending appeals through certified mail. Personalized appeals dramatically improved the effectiveness of such political appeals from 38% to 78%. (Dillman et al., 2008) Of course the internet, as Karpf (2012) has convincingly argued, further intensified personalized appeals through computational marketing techniques that targeted likely supporters thus rationalizing scarce campaign resources (Howard, 2005). Harnessing the audiences on political blog networks and discussion groups, mass personal appeals produced a revolution in fundraising supported by a vastly expanded system of micropayments enabled by new online payment systems (Wilcox, 2008). As a consequence, by January 2007, ‘ Obama’s campaign announced that it had raised more than $7.5 million over the internet in just 36 hours’. (Wilcox, 2008: 1)

The increase in personal appeals was made even more intimate in the 2020 Presidential election, with both presidential campaigns making substantial use of text messaging appeals directly to voters’ mobile phones. A research report published from the Center for Media Engagement at the University of Texas surveyed the role that mobile apps and texts play in personalizing political forms of misinformation. And while the majority of the report focused on the spread of false voting information and other forms of texted political disinformation, it also noted that such texts have started to link to digital wallet apps, making fundraising much more seamless, user friendly and automated. Political communications and consulting companies have likewise begun to produce personalized voter support cards for digital cell phone wallets (Glover et al., 2020).

While such studies point to the important role that technology plays in facilitating micropayments through personalized and customized fundraising appeals, they also lay the groundwork for understanding the circulation and sharing of factually questionable claims in fundraising appeals. Given that such appeals are often directed at self-identified partisans, previous contributors (Hassell and Monson, 2014) or ideologically committed voters, particularly after 2002, (La Raja and Wiltse, 2012), such financial appeals often focus on divisive ‘red meat’ issues such as abortion, immigration and gun control. Such personalized appeals can also include personal attacks, calling for funds to target the opinions or actions of politicians, public officials, and reporters. Miller and Krosnick (2004) concluded in an early fundraising study that pitches that included direct political threats produced the greatest amounts of financial contributions.

Oklobdzija (2017) has also noted that lawmakers have benefited financially online by posting extreme political positions during periods of heightened political tension such as budget impasses. In this historical context, it should come as no surprise that when falling behind Joe Biden’s fundraising efforts in the 2020 Presidential campaign, Donald Trump’s email fundraising appeals became increasingly more personal, urgent, and factually incorrect. Researchers have reported that Trump’s fundraising texts sent during the waning days of the 2020 presidential election became increasingly desperate in tone, again invoking very personal language to solicit contributions. One such fundraising appeal prior to election day for example read ‘Do you not care that Donald Trump is running for re-election? Donate now’. (Zarroli, 2020). A series of other fundraising appeals encouraged potential donors to help build a ‘Trump Army’ that could ‘fight off the Liberal MOB’ (Grady, 2020). During the month of November 2020 Trump reportedly sent over 500 emails to supporters, 400 of which sought financial contributions (Roche, 2020). However, as Trump intensified his fundraising efforts on the back of his attempts to undermine the legitimacy of the presidential election, the success of Democratic fundraising over the same period was widely attributed to the technique of ‘Fund-raging’, using strong partisan language to attack and target the behaviour and opinions of Trump. Thus, as one journalist wrote, ‘If 2008 was about hope and change for Democrats, 2020 is about anger and fear’. (Davis, 2020)

To summarize, while rigorous and interdisciplinary, the vast majority of major research studies of disinformation published over the last few years has exhibited a clear blind spot, a gap in understanding the role that finances play in funding the spread of disinformation. Such forms of communication are largely restricted to social media sites and other online spaces where users post questionable or incorrect information and conspiracy theories. However, as Swartz (2020) recently argued, ‘. . .what we’re seeing now is a shift from mass money media to social money media’. (p. 18). In short, money has become social media, with online groups forming around financial transactions. And the same can be said of fundraising. GoFundMe for example encourages its users to ‘create a successful crowdfunding fundraiser’ by building an online community through ‘Facebook, Twitter, hashtags, email, websites and blogs’.² Another prominent advice page on the crowdfunding platform lists seven strategies for ‘turn[ing] social media followers into donors’.³

We should caution however that while crowdfunding is well designed to facilitate viral forms of communication, this should not alone lead us to any conclusions about its role in amplifying disinformation. Rather, to better understand crowdfunding as a site of disinformation, this paper first sets out to determine, following the political fundraising literature, if pitches likewise incorporated personalized, and at times abusive language in their pitches for donations. Do campaigns include personal appeals or attacks? Do they target specific individuals, groups and communities? And are the campaigns themselves personal endeavors, that is are they launched solely by individuals, or by groups or larger associations? Comparing our two crowdfunding issues might also help to qualify how crowdfunding facilitates disinformation. Are ‘5G’ campaigns less civil and personal in tone compared to ‘election fraud’ campaigns? And what role – if any – does networked content play in affirming the claims made in networked pitches in campaigns? How might such content amplify forms of disinformation?

Methodologically speaking this paper deploys a ‘first-person’ or ‘interface’ based research perspective (Elmer, 2015; Marres and Gerlitz, 2016) that privileges the experiences of everyday technology users, over analyses of disaggregated data – information that is typically scraped from user interfaces. This perspective is influenced by Markham’s (1998) step-by-step methodical account of the choices taken, and functions enacted in online environments. Light et al. (2018) advanced Markham’s first-person approach by systematically following a user’s practices from first registering for an app and following its options and potential uses through to its deletion and removal. Our methodological approach, in short, invokes a pragmatic user-aligned perspective that highlights the common affordances and content presented on online platforms. We adopt this first-person perspective specifically to justify our small sample of campaigns and consequently the generalizability of our findings.

The first-person approach takes the view, following Grusin (2000) and countless user engagement studies, that the first page of results on a user’s computer screen are of immense strategic power. Rogers (2000) similarly refers to this screen space as the site of ‘preferred placement’. One industry study found that 95% of web traffic stems from the first page of results. (Leverage Marketing, 2017) It is for this reason that we decided to restrict our qualitative study to the first results displayed by the search function on the GoFundMe crowdfunding platform, a site that like many others algorithmically promotes ‘trending’ (i.e. controversial and attention seeking) content (Paust, 2021).

The study focuses on two distinct search terms that capture both an electoral form of disinformation and a political campaign around an issue of community concern. Taking into account the controversy surrounding the 2020 election, we settled on the search term ‘election fraud’ to study election-based campaigns and ‘5G’ in an effort to capture campaigns that question the health implications of the new wireless protocol. The term ‘fraud’ emerged as a catch all term widely used in the media to represent accusations in an election process was procedurally corrupted. The term’s legal meaning served as a cornerstone for legal appeals that followed the 2020 U.S. presidential election. (McEvoy, 2020) While the 5G issue was less sensational than claims of election fraud, numerous reports have emerged of groups organizing to fight – sometimes using destructive tactics – the installation of 5G towers in their communities. (Reichert, 2020) Some reports have also suggested that 5G, COVID-19 and election conspiracies have begun to merge online through QAnon sites. (Spring and Wendling, 2020)

On February 4, 2021 inputting the search terms ‘election fraud’ and ‘5G’ through the GoFundMe search tool produced six campaigns viewable on the first page of search results. The authors manually collected (cut and pasted) the title of each campaign, the URL, the pitch text, the name of the campaigner, the funds collected, and all URLs embedded in the campaign pitch directly from the platform interface.

The following section summarizes the findings from our six ‘election fraud’ and six ‘5G’ campaigns, beginning with the personalization of campaigns, use of uncivil language, crowdfunding targets, and lastly networked content used to supplement the pitch’s claims. But first let’s begin with some context on the campaigns. All 12 campaigns found through our two search queries reinforced the American and English-language bias of the GoFundMe platform, mirroring the findings of our previous study of COVID focused crowdfunding campaigns (Authors anonymized). Campaigns were based in the United States (8) Australia (2) England (1) and Spain (1). All six of the ‘election fraud’ tagged campaigns were launched from the United States. By comparison the 5G issue was decidedly more international, with campaigns from Australia (2), the United States (2), England (1), and Spain (1). The 5G campaigns netted a total of $101,776, compared to only $49,400 for the election fraud campaigns.

The campaign pitches, visually the most prominently framed content on GoFundMe campaign pages, represent the rationale for the campaigner’s financial appeal. We sought to determine if the personalization of financial appeals demonstrated in our political fundraising literature review were also commonly deployed by crowdfunding campaigns. By this measure, the two campaign samples were clearly distinguished by the nature of their appeals. Four of the six 5G campaigns made consistent appeals to a collective and community. The one Spanish language campaign exclaimed ‘we need to have “Patrons” who support our cause, which is that of the entire Spanish Society’.⁴ An American based campaign similarly promised that funds would ‘protect the entire US. . .our children, our environment and ourselves. . .’ Campaign pitches in other words often ended with an appeal to their perceived community. One campaign urged its readers – ‘together we can do this’, while another called for financial contributions to protect ‘our loved ones’. By contrast, the election fraud campaigns were dominated by personal appeals, with four of the six consistently pitching the needs of the individual campaigner, including one who was set to lose their home. These campaigns typically focused on funding the participation of an individual in a political process, such as attending a protest or conference. Another campaign named an individual that personally sought to address alleged vote counting irregularities as the beneficiary.

An unanticipated finding relating to the personalization (or not) of the campaign pitches, was the blurring of the role of the campaigner. Since the platform requires the campaigner to name the beneficiary of the funds collected in a dedicated field on the platform, we were able to collect this data for each campaign. We found that nine of the twelve campaigns were what we call proxy campaigns, that is campaigns launched for the benefit of third parties. Six campaigners sought funds for organizations, while three sought contributions for other individuals. While the individual campaigns merely noted ‘[name of campaigner] is organizing this fundraiser’ above the campaign pitch on the GoFundMe campaign page, the proxy campaigns typically noted ‘[name of campaigner] is organizing this fundraiser on behalf of [name of beneficiary]’. While some of these proxy campaigns included appeals from only one person, they also noted their past association with a beneficiary organization, while another campaign noted that funds were going to an organization that was headed by the campaigner.

In addition to analyzing the use of personal or social appeals in crowdfunding pitches, we also sought to determine if such text included uncivil or toxic language. To determine if such language was an element in our crowdfunding samples we slightly revised Tenove and Tworek’s (2020) framework to categorize uncivil social media posts to better account for the context of crowdfunding. In addition to accounting for pitches that altogether avoided posting uncivil language, we anticipated three levels of incivility, rising from the least to the most serious. Low incivility would include the use of hyperbole, dismissive or disrespectful language, for example, ‘this government is completely clueless’, while a middle range would exhibit direct insults, negative stereotypes, and offensive language such as ‘this politician is a [expletive] lazy idiot’. The highest level of incivility would include hateful language and direct threats.

We found a mix of civil and uncivil language in the 5G campaigns. While half the campaigns didn’t meet our threshold for incivility at all, three others made subtle or direct legal threats. One campaign claimed that politicians ‘. . .may face legal action under the jurisdiction of criminal law, public office laws or personal suits of civil action’. Given that the claim was qualified (‘may’), we categorized it as a medium level of incivility, however two other claims included much more explicit and direct threats and claims. One campaign for example claimed that ‘We are all being lied to by the wireless industry, FCC, etc. They are corrupt and need to be brought to court’, while the other suggested that the UK government was ‘. . .committing a Human Rights Crime against the population without their consent’.

While our second set of campaigns were already tagged ‘Election fraud’ none contained any high or medium levels of uncivil language. Four campaigns for example lacked any uncivil language. These campaigns instead used language that communicated trust and accuracy. One campaign even sought to downplay or at least qualify the claim of election fraud altogether, using scientific language: ‘This is a pattern that is consistent with a hypothesis of election fraud being perpetrated’. One other campaign added more hyperbolic language to the claim of election fraud, labeling it as ‘rampant’ and ‘blatant’.

Another complementary line of inquiry focused on the targeting of specific individuals, groups or organizations in crowdfunding pitches, again mirroring tactics deployed in political fundraising campaigns. An analysis of such targeting compliments our previous study of the tone of fundraising appeals, again getting to the question of how the fundraising pitch invokes a perceived threat that urgently needs funds to counter. By categorizing our campaigns by their named ‘targets’ we can also determine how such pitches frame the source of the perceived problem. Four of our 5G campaigns targeted ‘the government’. In addition to the general targeting of governments, each of the campaigns posted different secondary targets: the courts, another noting politicians, and another even ‘voters’: ‘I ask for support . . .so that I can wake up enough voters in time for them to make the right choice to select an MP who will stand up for their very right to life’. Only one campaign that targeted the government also included a private actor as a secondary target: ‘We are all being lied to by the wireless industry. . .They are corrupt and need to be brought to court’. And one other campaign simply listed the courts as their object of scorn and target for fundraising. Three election fraud campaigns by comparison targeted governments in their fundraising pitches. One of these campaigns also targeted the courts: ‘An election contest (challenge) was filed in the county court but dismissed as not timely by the Presiding Judge though the law is easily understood as to the deadlines’. Another campaign included a partisan target, the Democratic party, in addition to the “corporate media”: ‘Now I’m back to document the election fraud of 2020 and to document Bernie Sanders’ struggle against the Democratic Establishment’. No election fraud campaigns targeted the private sector. Two campaigns listed no targets in their fundraising pitch.

Lastly, while our study focused on questions of fundraising, we also contend that crowdfunding should be considered as a communicative form, one that might change the beliefs and behaviors of potential donors. To that end we questioned what other appeals campaigns explicitly listed in their pitches. Save one product-based campaign that sought investors for 5G safety clothing, all campaigns sought funds for either legal costs or various forms of political action. Half of our entire sample sought funds to launch or support legal battles, lawsuits or appeals, while the remainder sought to fund various political campaigns, including lobbying efforts, electoral campaigns, and costs associated with attending political events and conventions. One campaign sought funds to produce a documentary film. There was little variation in the two different issues, with both samples represented by three campaigns calling for legal funds. Both sets of campaigns also sought funds to support efforts at engaging with various levels of government.

Our last question focused on the role that networked content might play in amplifying disinformation. Not surprisingly, given the issues we chose to study, we found a range of networked content on crowdfunding pitches that spanned the spectrum of disinformation and misinformation. First, with regards to our 5G campaigns, only one campaign lacked any networked content, though its pitch offered a number of questionable and unsubstantiated claims about science and compared the rollout of 5G technology to Nazi human experimentation. All five other campaigns included robust links from their campaigns to other platforms and content. One campaign however only referred potential donors to their website that sold clothes that purported to protect against 5G signals. And while the site employs a typical commercial design, with links to a clothing shop and the organizer’s Facebook page promoting the clothing, it also includes a documentary film on 5G harms replete with links to a series of debunked conspiracies from chemtrails to antivaxxer sites. Thus while at first glance the campaign seems to focus largely on clothing, the associated website offered a plethora of disinformation on a wide range of topics.

The remaining four 5G crowdfunds share a number of similar characteristics found on their networked pitches. Principally, the linked content served to confirm the legitimacy of their claims by referring to likeminded campaigns, most prominently legal proceedings to stop 5G installations. All these campaigns linked to news content that reported on legal challenges to 5G initiatives, though only one campaign cited any substantive decision that successfully supported the claims that 5G signals may lead to health issues. This same campaign also twice linked to a purported United Nations policy against 5G technology. On closer inspection however the linked document is only a statement brought to the U.N. by an N.G.O. and not an adopted policy.⁵ These four campaigns also made common use of YouTube to again highlight proceedings against 5G interests, though once more there were no definitive legal decisions supporting claims about the health effects of 5G technology. The other use of YouTube video was self-referential, to highlight the crowdfunding pitch, the campaigner’s central website, or their own community based and legal efforts to stop 5G installations. Taken as a whole the networked pitches of these four campaigns serve to reinforce the belief that there is a legitimate, sustained, and unjust treatment of opponents of 5G. On the science of 5G however there is remarkably little linked content.

Compared to the 5G sample the ‘election fraud’ campaigns offered far fewer links to other networked content. Only two of the six campaigns included links. One campaign included a video appeal in the pitch, while the other included a link to a blog and another to a personal website, both authored by the crowdfund campaigner. In short, all three links in our entire election fraud sample exhibited self-referential content or what some researchers have dubbed ‘ego-networks’.⁶ One such linked crowdfund pitch for example stood out from all others, adopting a critical discourse commonly associated with Bernie Sanders political positions, and others who criticize the party’s corporatist ties. This campaign pitch linked exclusively to a series of short anonymously produced documentaries that speculate on the future of American politics if Bernie Sanders would have won the presidency. The other campaign that included linked content mirrored the 5G pitches where news content, blog posts, and other social media links simply referred to content produced by the crowdfund campaigner. The ‘election fraud’ pitches that lacked any hyperlinked content all obviously make claims that were not substantiated with any secondary sources. While the goal of the paper was not to verify every claim made in such crowdfunding campaigns, the pitches repeat a similar pattern of claiming an injustice based upon little to no evidence of wrongdoing. The other conclusion we draw from our brief analysis of networked pitches, from both samples and linked/unlinked pitches, is the prominence of the campaigner, in short the articulation of how the campaigner is seeking (funds) to right a perceived injustice.

What are the lessons learned then from these findings? First, what distinguishes the crowdfunding pitch as a potential site and form of disinformation? And second, how do our findings frame the crowdfunding platform, and its attendant practices – most notably the networked pitch – as a potential site and source of disinformation?

What most clearly distinguished crowdfunding as a form of disinformation was the preponderance of appeals for funds to initiate or otherwise support ongoing legal proceedings. Both of our query samples focused strongly on perceived injustices of court decisions. While there were political and partisan appeals in both samples they were broader in scope and focus, that is largely used as background information. The specific financial rationale in crowdfunding pitches by comparison were explicitly tied to supporting the costs of legal action. Such findings suggest that disinformation campaigns are obviously not restricted to political events, news outlets or social media sites. Rather disinformation campaigns are being amplified as legal matters beyond the political sphere, above and beyond partisan political belief systems and democratic processes.

The targeting of the courts by crowdfunding campaigners might further prolong and extend the spread of disinformation, given that we found evidence that campaigns sought to use their appearance in court as a form of political legitimization. In other words, the courts were to some degree recognizing or at least considering their claims. Even more concerning is that campaigns threatened court action as part of their crowdfunding pitches. In short, crowdfunding disinformation is weaponizing the court system against those who call out their lies and incorrect information. Future research could determine if such threats are – like in other areas of public speech – having a chilling effect on those seeking to address the disinformation campaigners, or the extent upon which such legal efforts are impacting the publicly funded court system.

Some of our campaigns noted that the courts were the site of last resort, the last place where they could seek justice. Consequently, as courts continue to rule against purveyors of disinformation one might conclude that some disinformation actors could become increasingly more radical and socially disruptive in their goals and appeals. At this stage however our findings however do not support this conclusion. Following the literature on political fundraising that highlighted more personal, vitriolic and urgent appeals for political donations, we were surprised to find that our crowdfunds were relatively civil. The election fraud campaigns were overwhelmingly civil in tone, with the personalized pitches suggesting that the crowdfunds were part of an individualized (often self-aggrandizing) campaign to expose the truth. And on the whole the election fraud campaign pitches were much decidedly thinner in content – they contained far fewer links to other content. The campaigns in other words seemed to be much less ambitious and focused in their goals. Such findings might suggest that they are being deployed more as expressions of political frustration or outlets for short term grievances, compared to the more established, networked and articulate claims of the community framed 5G campaigns.

Turning to the networked content collected on our crowdfunding campaigns. The expansive use of cited and linked reports, studies and organizations among the 5G campaigns suggests a deeper engagement with matters of science, political processes, and community engagement. Such networked campaigns also highlight the opportunities for users to read not only fundraising pitches, but also the extensive set of resources that we found accompanying them. Furthermore, one brief but noteworthy finding we did not fully discuss above may be of relevance herein, that the only campaigns that asked for action beyond donations were in the 5G samples, two campaigns that urged potential donors to also ‘educate yourself’ or read listed links to 5G related content on other websites and platforms. Such forms of ‘connective action’ (Bennett and Segerberg, 2012) were unique within our sample and distinguish the approach that the campaigners brought to the two issues.

The most notable findings emanating from our study of the networking of pitches however concerned the initiators of the campaign and not explicitly linked content. That said, we might consider the campaign initiators as the crucial node in a crowdfunding network that, at least in the 5G sample, further extends out to other potential forms of disinformation. We were again very surprised to find that a large proportion of our campaigns – from both issues – were proxies, that is were being launched for the benefit of other parties. This raises a series of questions about not only the financial and political regulation of disinformation crowdfunding, but also its potential use as a distributed form of fundraising. It is conceivable that only one individual or organization may be launching and running multiple campaigns in support of one or more dubious political and scientific claims. Conversely like the Russian crowdsourcing study noted in our literature review, some campaigners may be leveraging the network of popular disinformation actors to collect funds for purposes beyond those of their proxy. In either case, there is potential for forms of disinformation to proliferate.

While our central goal was not to specifically quantify and verify the extent of disinformation on crowdfunding sites like GoFundMe, there were clear indications of disinformation in our campaigns, including a plethora of claims and references to documents that lacked credible sources, were hosted by unknown organizations with no contact information, or were factually incorrect (e.g. claims of an anti-5G UN policy that was merely a statement brought forward by an NGO). Future research that traces the circulation of documents on crowdfunding platforms would help to quantify and qualify the impact of specific texts, claims, discourses and mis/disinformation actors. Future research could also complicate or otherwise contribute to our first persona approach to users accessing potential forms of disinformation. Where do users first access such content? Directly from discussion forums, or do they predominantly read such information via social media, search engines, or search functions on platforms such as GoFundMe?

Furthermore, given our sample size we also do not make any firm representative claims about the relationship between the pitch and the amount of funds collected, that is the success of the financial appeal. Such a question would require a much larger study that compared issues, pitches, and of course funds collected. Our one contribution to future studies of crowdfunding success though is the marked difference in the amount collected by our two samples. Again, while working with admittedly small samples, the 5G campaigns notably collected twice the contributions of ‘election fraud’ campaigns. Such a difference might also be explained by the much more liberal use of links to other sources of information and videos. In short, the 5G campaigns made much greater use of networked content that the platform itself suggests leads to greater user engagement and donations. We could also speculate that the prevalence of social appeals in the 5G campaigns, compared to the highly individualized pitches from ‘election fraud’ campaigners, might also contribute to donors relating to the community impact of the 5G issue.

While the expressed goal of crowdfunding platforms is to facilitate fundraising, we have also demonstrated herein that the pitch of the campaign also contains considerable information that may contribute to the spread of incorrect or misleading information on important matters of public concern, science, public policy and trust in institutions and experts. Such campaigns are also integrated within preexisting political networks, both online and offline. And the platforms themselves as previously noted offer a robust set of tools to easily share and otherwise spread the campaign appeals across multiple platforms. Just considering the range of options to spread information from GoFundMe, we could argue that the crowdfunding platform is more fully, infrastructurally vested in online networks than most if not all social media sites or other platforms that tend to encourage or even restrict user interactions and engagements to their own platform or co-owned properties. The crowdfund platform by design craves cross platform attention and engagements for their campaigns. As a consequence, we should carefully consider at least infrastructurally speaking, the crowdfunding platform as a potential super spreader of disinformation.

While the crowdfunding platform seeks links across the internet, it also places the campaigner’s pitch at the center of the crowdfunding proposition. The pitch though clearly diverges from those discussed in the political fundraising literature, where political leaders, amplified by new personalized technologies and techniques call for contributions largely from previously identified partisans. Our crowdfunding campaigns mirrored some of the language and tactics of such fundraising appeals, in their tone and sense of urgency. But crowdfunding campaigns also appealed to individual grievances and frustrations with little vitriol, as noted they seemed more an afterthought than a strategic or emotionally charged intervention. We might thus consider, at least these election fraud campaigns as sites of personal negotiation, as spaces where opinions are floated – ideations that nonetheless may be rife with disinformation claims and conspiracies.

While our empirical goals in this paper were modest by design, we believe the qualitative study of crowdfunding pitches has a myriad of contributions to existing arguments in the contemporary disinformation literature. Our first-person approach also importantly recognizes the importance of user practices, specifically the manner in which they seek and consume online disinformation.

First, our findings suggest that as agnostic properties, that is platforms that do not seek to enclose their users, crowdfunding platforms are among the most networked and cross-platform enabled sites on the internet. This should come as no surprise given that such platforms are, unlike the vast majority of social media sites, explicitly designed to conduct campaigns. The attention economy, highlighted by Marwick and Lewis (2017), is thus fully optimized by crowdfunding platforms like GoFundMe as they seek to cast a wide net to attract potential donors. There was also some evidence in our study to suggest that campaigners were actively recommending and encouraging users to read and further investigate unverified and misrepresented information on 5G risks.

Perhaps the most important contribution to the literature from our study though is the role of the individual campaigner in the context of broader concerns over the spread of disinformation. We are not suggesting that campaigners should be compared to Steve Bannon and others who have leveraged vitriolic and hateful speech to solicit funds on GoFundMe. But we propose that a larger study would very likely show, based upon our small sample, that the aggregate amount of funds collected to support court challenges, appeals, and other forms of political organization, are substantial and serve as a model for others to emulate.

While we have been able to enumerate some elements of the ‘weaponizing’ of crowdfunding, including the targeting of certain sectors and legal actors, there remains much more to do. A study of the sharing of disinformation documents and resources from campaign pitches is among the most urgent and would be greatly assisted by a digital methods approach that could deploy computational tools. Comparative studies of other crowdfunding platforms and issues are also obviously needed, and again with larger samples and/or more focused qualitative goals. But based on this early study there can be little argument that pitching for money can not only provide material support to those spreading disinformation, it can also amplify and further legitimate preexisting discourses and sources of disinformation.