Nº 68 | OCTUBRE 2025

ISSN: 1139-1979 | E-ISSN: 1988-5733

Cómo citar: Martins, A.V. (2025). Origin and format of disinformation against Brazilian indigenous peoples. Ámbitos. Revista Internacional de Comunicación, (68),110-126. https://doi.org/10.12795/Ambitos.2025.i68.06

Origin and format of disinformation against Brazilian indigenous peoples

Origen y formato de la desinformación contra las poblaciones indígenas brasileñas

Allysson Viana Martins

Federal University of Rondônia, Brazil | Campus – BR 364, Km 9,5, Porto Velho, Rondônia

0000-0001-7805-6512 | allyssonviana@unir.br

Recepción 19/01/2025 · Aceptación 31/07/2025 · Publicación 15/10/2025

Abstract

Brazilian fact-checking agencies primarily focused on political statements rather than on the spread of fake news on social media platforms until 2018, when the focus shifted. This article aims to identify the channels through which false information targeting a vulnerable group, such as indigenous peoples, circulates and the media formats used in its dissemination. The Lupa fact-checking agency conducted, between 2018 and 2023, 36 verifications related to Brazilian Indigenous communities, which we evaluate based on the “definition” aspects of media framing methodology. Regarding the origin of false information, social media has proven to be a priority area (89%) for its dissemination, changing the historical way agencies operate and even leading them to partner with companies that use these technologies. In terms of formats, Facebook text and WhatsApp video were the most prevalent, either alone or in conjunction with other resources. Therefore, social media requires the attention not only of agencies but also of those who circulate information in these spaces, especially in text and video formats.

Keywords: disinformation, fact-checking agencies, indigenous peoples, social media, Internet.

Resumen

Hasta 2018, las agencias brasileñas de verificación de datos se centraban más en las declaraciones de los políticos que en la circulación de noticias falsas en las redes sociales. Desde entonces, ese escenario ha cambiado y este artículo busca identificar dónde circulan informaciones falsas sobre un grupo vulnerable, como las poblaciones indígenas, y qué formatos de medios se utilizaron. La agencia Lupa hizo, de 2018 a 2023, 36 verificaciones centradas en los pueblos indígenas brasileños, que evaluamos con base en los aspectos de “definición” de la metodología de encuadre de los medios. En cuanto al origen de las informaciones falsas, las redes sociales han demostrado ser un área prioritaria (89%) para su difusión, cambiando la forma de actuar históricamente de las agencias e incluso llevándolas a asociarse con empresas de esas tecnologías. En el caso de los formatos, el Texto en Facebook y el Vídeo en WhatsApp fueron los más presentes, solos o acompañados de otros recursos. Por lo tanto, las redes sociales requieren la atención no sólo de las agencias, sino también de quienes circulan información en estos espacios, especialmente en texto y vídeo.

Palabras clave: desinformación, agencias de verificación, población indígena, redes sociales, internet.

1. Introduction

The term ‘fake news’ gained widespread popularity and notoriety during the 2016 U.S. presidential election, which featured candidates Donald Trump and Hillary Clinton. While Clinton used the term to highlight a contemporary form of information disorder, Trump used it as a weapon to discredit unfavourable news coverage, regardless of its factual basis, when it contained only a few occasional errors (false news) (Carvalho et al., 2019; Cazetta & Reis, 2019; Derakhshan & Wardle, 2017; Eichler et al., 2018; Fernandes et al., 2019; Meneses, 2018; Spinelli & Santos, 2018). By fostering public distrust in journalism as an institution, Trump contributed to a communication environment conducive to disinformation. This tactic helped normalise the use of the term ‘fake news’ not only to describe fake content but also to undermine legitimate reporting. “While it is currently used to describe false stories spreading on social media, fake news has also been invoked to discredit some news organisations’ critical reporting” (Tandoc Junior et al., 2017, p. 138).This contemporary process of information disorder should be understood as a novel phenomenon, distinct from traditional rumours or slander, as in Cazetta & Reis (2019) and Moreno & Moutinho (2018). Beyond the reach of digital networks, there are ideological and financial foundations that cause false information to be produced and spread at an industrial pace and logic. “The novelty is not in the process itself, but in how it has been practiced and disseminated, causing tensions and singular phenomena by agents who work in favour of private interests of an actor or a group to achieve specific ends” (Silva et al., 2019, pp. 416-417).

Here, it is obviously not about denying that in previous periods, especially in power struggles such as wars and political campaigns, the spread of false information towards disinformation in a deliberate and planned manner was not common; Neither disagree that these problematic times of social upheaval make the belief in lies spread easier. However, we argue that the internet has enabled the dissemination of false information on an unprecedented scale, contributing to a unique form of informational disorder, once digital media makes the production easier, makes the publication cheaper, allows greater editing and manipulation, increases flow and automates dissemination.

Fake news, then, is a modern phenomenon in which false information is presented as credible and disseminated through digital media, particularly social networks, with the strategic goal of misleading audiences. These messages may imitate journalistic formats to increase their perceived legitimacy. This imitation serves to confuse those who receive false information and give some credibility to that content, thus facilitating disinformation and its spread easier. Schudson & Zelizer (2017) also recognize this distinction caused by digital media.

Whoever plans it knows it is a lie. And he or she only plans it because it is a lie. It only exists because it is false. It may not be completely false, as one of the techniques used to make fake news credible, and thus reach more consumers, is mixing elements (names, places, previous facts, photos, etc.) with lies. Rumours and –gossip have always existed, but they were rarely dignified by journalism. Rumours and gossip appear today, side by side with real news, in our social media feeds. (Meneses, 2018, p. 47).

They are also part of a heterogeneous ecosystem of information disorder, segmented into a triad, which we may translate from three different expressions in English, such as ‘error’, ‘mistake’ or ‘incorrect’ (misinformation); ‘fraud’ or ‘lie’ (disinformation); and ‘decontextualisation’ or ‘bias (malinformation) (Cazetta & Reis, 2019; Derakhshan & Wardle, 2017; Gruszynski et al., 2020; Meneses, 2018; Santos & Maurer, 2020; Wang, 2020). Although the first two actually deal with the publication of false information, in misinformation, there is no intention on the part of the disseminator to harm others. Malinformation involves the dissemination of true information, but biassed or out of context, thus causing informational disorder. That is to say, although they are facts, there is a negative intention. Disinformation, in turn, conveys lies intended to confuse.

When considering the reality of Brazilian indigenous peoples, Martins (2024, 2025) has highlighted the silencing to which they have been subjected over the past five centuries as part of the process of informational disorder. Silence is understood not as a total absence of information, but rather as a lack of information that opens the door to the spread of stereotypes and prejudices.

Since the explosion of the fake news phenomenon, especially after 2016, fact-checking agencies have expanded their focus beyond politicians’ statements, placing more emphasis to disinformation about broader and more diverse topics. Since the verification of politicians (Carvalho et al., 2019), agencies have broadened their focus and work with statements from all kinds of public figures, transmitted through advisory services or journalistic vehicles, and with fraudulent content circulating on social media and in society. Lupa, for example, has taken on this role of fact-checking more vigorously since 2018.

According to Dourado (2019), Brazilian fact-checking agencies generally focused hitherto more on politicians’ statements in their own media or traditional media than on the circulation of fake news. If they started to evaluate the speech of politicians and public figures nowadays, these agencies would appear as a “vaccine” against fake news and join other combat strategies, such as exclusion of content and demonetization of disseminators by companies like Facebook and Google (Eichler et al., 2018; Gruszynski et al., 2020; Silva & Melo, 2020; Spinelli & Santos, 2018), in addition to the establishment of laws and media literacy.

The flow of communication with digital media has reached proportions different from those of traditional mass media, being now broader and more heterogeneous, in addition to including a private messaging application. Until 2023 (Dourado, 2024), while Facebook, with approximately 110 million monthly users in the country, allows a public scrutiny of its messages, WhatsApp, with over 170 million users in Brazil, enables private sharing, making it difficult to combat that disinformation, especially when it comes from people they are already connected with.

This article is part of a broader research project aimed at identifying patterns in fake news and fact-checking – related to the disinformation spread about indigenous peoples in Brazil and evaluated by the digital fact-checking agency Lupa, also analyzing what the indigenous peoples involved in the lie said, in the case of public demonstrations or in fact-checks. In this paper, the emphasis is on fake news, with the intention of identifying where they circulate and which media formats were used for their propagation. To understand the schemes for interpreting reality in the properties and operations of the spread of false information about Brazilian native peoples, we use the media framing methodology, based on Entman (1993), Vimieiro & Dantas (2009) and Vimieiro & Maia (2011), focusing on the origin and formats.

The concept of media framing considers that reality and events are perceived through specific interpretation schemes developed by those who produce the information, even though they are also affected by other frames. It is also observed in the subtext of the message when “some aspects of a perceived reality are selected and made more salient in a text, in such a way that a definition of a particular problem, a causal interpretation, a moral evaluation, and/or a recommendation for treatment of the described item is promoted” (Entman, 1993, p. 52).

The more detailed evaluation of the material was conducted based on the first element of media framing, among the four proposed by Entman (1993): defining problems – the first aspects related to the agent; diagnosing causes – the first impressions and presentations of those who receive the agent’s cause; making moral judgments – the effects and impacts of the initial agents’ actions; and suggesting remedies – justifications and ways to solve problems. Matthes & Kohring (2008), who had proposed a methodology based on these four aspects, state that the definition of a particular problem includes the subtopic in question and who the most important social actor is.

This means that we are looking for answers about fake news itself, not yet about the agency’s strategies to combat them, that is, where they come from and which media format they use. In general, we figure out the day the lie was spread, what subject it addresses, where it comes from and which media format it uses – aspects related to definition. Subsequently, we focus on the agency´s work in more detail, evaluating the labels adopted and the presentation of the lie (interpretation); the sources used and how they were accessed to debunk (dis)information, as well as how they reached the lie (evaluation); and finally, the narrative structure, the media resources – for verification, and association with another similar verification (recommendation). In this article, we work with aspects related to definition (origin and formats).

2. Preliminary data

Lupa is one of Brazil’s leading fact-checking agencies. Although it operates independently, its societal reach has expanded through partnerships with the Meta Group, the circulation of its content in traditional media, and its involvement in media literacy through courses and training programmes (Gehlen, 2018). This context above led us to observe that it started to dedicate the checks about the native people upon the arrival of the far-right to power in 2018. In Brazil, even prior to his election in 2018, Jair Bolsonaro had already made false statements about indigenous peoples, most notably the Yanomami, who later suffered a humanitarian crisis under his administration, marked by famine and forced displacement that led to mass deaths.

The silence of the indigenous peoples – aligned with the possible lies spread, which we intend to investigate from the Lupa fact-checking agency – during his government was so evidently planned that the Ministry of Health ordered a Public Health Emergency of National Importance (ESPIN) immediately 20 days after Bolsonaro left the presidency, according to the ordinance GM/MS No. 28 of January 20th, 2023. This is particularly evident when considering that of the 36 evaluations conducted between 2018 and 2023, 14 were published in 2023 alone, with a greater commitment to lies about the indigenous population, surpassing the number of the previous 3 years, when the agencies’ focus was on disinformation about the –COVID-19 pandemic.

In 2021, at the height of both the COVID-19 pandemic and the accompanying infodemic (Martins & Teixeira, 2023, 2025; Teixeira & Martins, 2020, 2022), indigenous peoples from different ethnicities, from São Paulo to Pará, expressed fear and hesitation towards the coronavirus vaccine. This fear did not only arise from cultural and contextual factors but also from the pervasive disinformation surrounding the topic (Cruz & Barbosa, 2021). In some localities, the refusal even resulted in the return of vaccines, which were intended for other population groups considered a priority at that time.

In response to the harmful effects of disinformation within their communities, several Indigenous groups undertook efforts to promote media literacy and debunk false information. One of these initiatives was led by a group of young indigenous journalists who began disseminating fact-checks both in person and through WhatsApp after discovering that members of their community were refusing to get vaccinated due to the belief that the vaccine contained a serpent and would turn recipients into Satan (Mansani, 2022). Other groups, facing similar challenges in vaccine acceptance, produced podcasts and organized workshops (Borges et al., 2022), while some focused on promoting their cultural narratives and traditions through social media as a way to counter the spread of lies (Macário, 2023). In all the mentioned cases, WhatsApp emerged as the main channel for the circulation of fake news.

More commonly, however, false information does not directly affect indigenous peoples by convincing them of its veracity but rather targets them as subjects of disinformation. In 2022, groups who supporting thenpresident and re-election candidate Jair Bolsonaro, spread lies about indigenous lands in Rio Grande do Sul (Marko et al., 2022). The fake news focused on the increase in the demarcation of territories and the expropriation of occupied lands in favour of native peoples, in case the candidate Lula became president.

In the first month of 2023, the beginning of President Lula’s third term, the Ministry of Justice and Public Security had to deny that the Yanomami population, who suffered due to the actions of illegal miners and the lack of public policies from the previous government, had any connection to the Venezuelan indigenous communities who arrived in Brazil as refugees (MJSP, 2023). The federal deputy Luiz Ovando, from Mato Grosso do Sul, even claimed that these Brazilian indigenous people were actually Venezuelan (Moreira, 2023; Soares, 2023). And it was only with the end of Jair Bolsonaro’s government that the Yanomami people were considered to be in a humanitarian crisis, due to their extreme malnutrition situations.

Towards the end of 2023, the Federal Prosecution Office of Pará requested the dismissal of the mayor of São Félix do Xingu, João Cléber, for spreading lies that the operation to remove invaders from an indigenous land was stopped and the intruders would receive compensation, as they had been there since the 2000s (Madeiro, 2023). Both pieces of information were not true, and the mayor is even accused of contributing to illegal occupation.

At the Lupa agency, from 2016 to 2023, there were 149 materials published containing the terms “indigenous person” and “indigenous peoples”, with the peak of publications in 2023, totaling 44. Even so, of the publications, 103 are fact checks with evaluation labels, with the other 46 consist of podcasts, interviews, reports, op-eds, comparisons or internal balance sheets. Of these 46 general productions, only 10 directly address issues related to indigenous peoples: 2 in 2019, 1 in 2021 and 7 in 2023, after Lula’s election as President of the Republic and the creation of the Ministry of Indigenous Peoples. The 36 other publications only have a brief mention of the two words aforesaid, but do not directly about the native people.

Regarding the 103 checks, our analysis corpus only comprises 36, as they directly address issues concerning the indigenous population, while the 67 other evaluations only mention the two keywords from our search. According to Figure 1, it is possible to note that in 2020, in the first year of the pandemic, the checks exactly about the topic decreased by almost 70%, compared to the previous year, when COVID-19 was not present in the world. And the amount in 2019 was only surpassed in 2023, the year of President Lula’s inauguration, the creation of the Ministry of Indigenous Peoples the World Health Organization’s declaration of the end of the COVID-19 pandemic, a topic that even accounted for half of the checks in agencies. This demonstrates how the indigenous issue has become more evident again, both for digital fact-checking agencies, like Lupa, and for spreaders of fake news.

Figure 1

Frequencies of information and checks on indigenous peoples in Lupa

Lupa began producing more content about indigenous peoples since the election of then-presidential candidate Jair Bolsonaro, with checks and posts based on what he and his followers were saying. However, it is clear that there is now a greater dedication to indigenous peoples than before, when we compare the frequency of publications and verifications from 2016 to 2022 with those from a single year of President Lula’s government, in 2023.

The absence of public policies in the Bolsonaro’s government, which almost resulted in the genocide and extermination of the Yanomami people, was likewise evident in the neglect of information dissemination and fact-checking related to Indigenous peoples; as in just one year (2023) of Lula’s government is equivalent to the last 3 of the previous administration. That is to say, even though the speech by far-right politicians and supporters has brought the subject and the lies about indigenous peoples to Lupa’s attention, this movement is strengthened by the current government by creating the Ministry of Indigenous Peoples and expanding attention around this population.

Thus, we may say that the Lupa agency evaluated, from 2018 to 2023, 36 information checks focused on native peoples, of which 33 (92%) present the label “False”, and, of these, six shared other labels, because there are other pieces of information which were evaluated, including “True” and “True, but...”. The only 3 (8%) analyses which do not contain the false label are placed as “True, but...”, though they contribute to the broader disinformation process (Martins, 2024). They address: (1) Jair Bolsonaro’s in 1998 praise for the US military for decimating indigenous peoples and not having problems with land demarcation (2018); (2) the death of a chief after receiving the vaccine, although his heart failure has no relation to the dose (2021); and (3) a 2021 video of an indigenous protest that was taken out of context to make it seem like a protest against Lula’s victory over Bolsonaro in the 2022 presidential elections (2022).

3. Origin and formats of fake news

The routine practices of fact-checking agencies can be summarized in three stages: retrieving digital traces, analyzing and verifying the facts, and presenting the results of the check (Silva et al., 2019). In other words, professionals investigate – especially data –, produce – considering links and multimedia – and edit – especially on defining labels and content distribution – their material carefully. This article specifically focuses on identifying the sources from which the fact-checking agency Lupa retrieved the disinformation about indigenous peoples they sought to evaluate, that is, in which spaces they circulated and what their formats were.

In this text, lies were evaluated in two aspects: 1) Origin of fake news, observing whether it comes from traditional journalistic media (like websites, TV, radio and printed media), websites and blogs or social media (specifying whenever possible); (2) Format in which the lie circulated, whether in text, image, audio, video or their various combinations. Fernandes et al. (2019) realized that almost 30 % of the material evaluated by the agencies comes from edited photos and videos which circulate on the internet. In Europe, 17% have as target, mostly (90%), statements by politicians in press conferences, journalistic articles or social media (Graves & Cherubini, 2016).

In relation to the origins of the 36 fake news about indigenous peoples analyzed by the Lupa, we noticed that 89% (32) came from digital social networks, of which 26 (72%) from Facebook and 6 (17 %) from WhatsApp. Even so, almost all of them derived from this network (4 of 6) appeared in 2023, whereas earlier instances of disinformation were predominantly found on Facebook. If we consider only this year, there were nine originating from Facebook and four from WhatsApp, proving the importance that this app has been gaining for sharing false information in our country, or, at least, the focus that Lupa has been giving it. The significance of social media as a source of disinformation is so substantial that fact-checking agencies typically disclose the platform of origin at the very beginning of their evaluations.

Only four cases (11%) did not originate from these (cyber) spaces; three came from the very journalism/press itself and one from a session of the National Congress. This in-person nature, spread only in text, was carried out by the Parliamentary Commission of Inquiry (CPI) on COVID-19 in 2021, an investigative committee of the Federal Senate chaired by Senator Omar Aziz to investigate alleged infractions in the federal government during the pandemic. “In the CPI, the senators once again make false statements about the pandemic,” assessed the speech of several politicians, two of them with statements, in this case true, about indigenous peoples. Fabiano Contarato said that “More than 1200 indigenous people lost their lives “[in the pandemic]” – and until then, there had been 1222 deaths – and Flávio Bolsonaro declared that “the indigenous peoples living in villages, more than 80% of them being adult population, are already completely immunized” – and until that moment, 81% of the adult indigenous people were vaccinated; although this data does not account for 42% of the indigenous population, who live in the city or in demarcated areas.

These four cases were the only fact-checks not prompted by user submissions on social media but rather initiated at the discretion of the Lupa editorial team. This means that all 26 checks that originated from digital social networks were carried out at the request of network users. This may be due to Lupa’s business relationship with the Meta Group since May 2018 to analyze the content circulating on its networks, such as Facebook (Lupa [...], 2018). A little more than two years later (Lupa [...], 2020), in October 2020, it launched an automated channel on WhatsApp in partnership with the same group to facilitate communication with the public, speeding up the verifications and the survey of fake news. Of these four, only one, derived from the press, had video as the only broadcast format, while the other three had only text, without any other resources.

In terms of media formats, text is the most prevalent, appearing in 32 of the 36 fake news cases (89%). Among these, 6 were purely textual, 10 were accompanied by images, and 16 were combined with video. However, in only one instance did the image merely serve an illustrative function, without adding information to the accompanying text (see Figure 2). In the case titled “It is true that Bolsonaro praised the North American cavalry for decimating their indigenous peoples”, the false information is in the text, placed in image only to present an art/card with a striking red face of Jair Bolsonaro, giving a statement in a contextualized manner; in fact, this case is one of the few lies about indigenous peoples that attacks someone from the far-right (4 out of 36), suggesting that the misinformation was probably not produced by individuals aligned with Bolsonaro’s political group.

The first two lies derived from the press (2018 and 2019) were spread only with text. In “Bolsonaro and indigenous peoples: the president-Elect’s mistakes in speaking about Funai and reserves”, Lupa selected statements by Jair Bolsonaro circulating about indigenous lands, with lies such as the number of indigenous peoples who live in the Yanomami reserve – an ethnic group who suffered a humanitarian crisis because of his government –with the former president saying there were only 9,000, when in reality, it extended far beyond 25,000 t. Regarding this and other statements, the agency reports that “when contacted, Bolsonaro did not respond.” In “on TV, Bolsonaro makes mistakes about fuels, indigenous reserves, and the 2005 referendum,” speeches by the politician were selected during an interview with journalist José Luiz Datena on the TV program Brasil Urgente, where he repeated the aforementioned lie about Yanomami people.

Figure 2

Illustrative image, without adding information to the malinformation in the text

The video is the only other format to appear independently, occurring in four instances (20 in total, 56%). Notably, the video format gained particular prominence starting in 2023. Between February and December of that year, video was present in all 11 analyzed disinformation cases, 10 of which also included text. However, the full video content of the disinformation was available in only one case, which originated in the press in 2023. This is because, when it originates from social media, Lupa always puts its print, not the video. However, in “Lira gets data wrong about indigenous peoples when defending time frame in Roda Viva,” when it was a lie during an interview with the then president of the Chamber, Arthur Lira, on the television program Roda Viva, the agency presented the video of the false information in in its entirety, claiming that the indigenous population of the country corresponds to 0.02%, when in reality it is 0.81%, a difference of 1.5 million people.

There is a clear correlation between video content and WhatsApp (see Figure 3). All six lies (100%) which circulated on this social media were in this format, and twice alone (1/3), without text. This indicates that video, rather than text, was the dominant format for fake news circulating on WhatsApp. The two cases in which video appears alone are the only two lies evaluated which originated from WhatsApp before 2023. In “video from camp in Brasília is not a protest sit-in on September 7th” (2021), the fake news actually presents footage from the Struggle for Life movement, from the Articulation of Indigenous Peoples of Brazil (Apib), who fought against the approval of the time frame in the Supreme Federal Court (STF). The video falsely suggested that the indigenous peoples were expressing support for then-President Bolsonaro, with captions such as, “and they say that indigenous peoples do not like Bolsonaro.”

Figure 3

Disinformation shared on WhatsApp in the most used formats, Text and Video

In “Video with ‘truths about Brazil in the environment’ has false and decontextualized data” (2022), the disinformation originated from a video shared by Jair Bolsonaro himself on X (formerly Twitter), but it circulated primarily through WhatsApp and was flagged for verification by Lupa through this platform. This led the agency to reach out to the politician once again: “The report contacted the president and will update this verification in case of a response.” Among the lies spread are the following: the Brazilian Amazon has 84% of its territory fully preserved, when in fact, 62% is in Protected Areas; and 66% of the entire national territory is completely preserved, when this percentage concerns native vegetation, but only 29% is protected against deforestation in Environmental Protection Areas, Conservation Units and Indigenous Lands.

In contrast, on Facebook, only one case involved just video. The remaining 25 cases included text – 10 accompanied by image and 12 by video. In “it is false that the NGO of Chief Raoni sent gold bars to France”, the lie is only in the video showing white people receiving gold bars as if they were given by the Brazilian indigenous leader; however, they come from a shipment that arrived from Ghana inspected by Italians. The fake news had already been shared in other countries, such as Mali, and checked by French agencies. Of the three which only contain text, with no other format, one of them has a series of lies involving Anita Roddick, founder of The Body Shop company, who supposedly met with Chief Raoni, being deceived and financially extorted, with her donations being used for smuggling. None of this ever happened.

The two other checks only with text were supposed news which circulated on Facebook and were based on news websites, highlighting the paradoxical role of journalism, which, despite being discredited by disinformation agents, is often imitated to give fake content an appearance of credibility. “The death of the chief in Acre is not related to the COVID-19 vaccination” refers to the death of an indigenous leader who has nothing to do with the CoronaVac vaccine he received but rather with his over ten-year battle against diabetes, as well as hypertension and heart failure. In “It is false that the entity which reported malnutrition among the Yanomami embezzled R$ 33 million,” organizations with different, albeit similar, names spread the untruth.

Across the analyzed cases, Facebook and WhatsApp emerged as the dominant platforms for the dissemination of disinformation. Text was the most common format on Facebook, while video predominated on WhatsApp. These findings suggest platform-specific patterns in the dissemination of fake news about indigenous peoples.

4. Social media in focus

With the spread of information on the internet, especially on social media, disinformation has proven to be a social phenomenon which needs to be combated. Its consequences go beyond distorting public perceptions of reality; disinformation can also incite harmful actions, both physical and virtual, perpetrated by individuals influenced by fake narratives. Technology companies themselves have responded through various measures, including content removal, user notifications, and demonetization of disinformation. These efforts have been complemented by legal frameworks and media literacy initiatives aimed at addressing the still ambiguous phenomenon of fake news.

The challenge is also being addressed through ongoing fact-checking efforts, carried out by agencies that have proliferated in digital environments, particularly on websites, some of which are affiliated with traditional news organisations. Fact-checking agencies play a crucial role in curbing disinformation on social media. They are especially significant in reshaping the mediation processes and cognitive filters that operate within digital platforms, and in the long term, they contribute to user education by promoting skills to assess the veracity of online content (Valentim, 2019).

It is also important to note that fake news does not always attempt to imitate journalistic formats or disguise themselves as professional reports (Carvalho et al., 2019; Fernandes et al., 2019); that is, fake news does not necessarily adhere to “a journalistic textual structure (title, lead and body of the text) and respect peculiar characteristics (short, clear, informative and impersonal text)”, with exaggerated headlines, attractive topics, and without legitimate sources or channels (Silva et al., 2019, p. 416). The spread of disinformation on social media, mainly on WhatsApp, does not necessarily rely on a news structure but rather on eye-catching content that facilitates its rapid propagation.

The problem is that WhatsApp is one of the main channels for spreading disinformation, especially in Brazil. As it is a private messaging application, the controlling for correction is far more difficult to reach for those who receive and transmit false content. WhatsApp “causes political discussion to migrate to private environments, where there is no content control, relying solely on the user’s task to identify whether some information is true or false” (Santos & Maurer, 2020, p. 4).

Gruszynski et al. (2020) also talk about fake news as imitators of news, although they emphasize the spread of disinformation in testimonial format later – not journalistic production –which supposedly would not have been published in the traditional media. Fake news does not need to resemble journalistic articles from a website with this appearance; that is, they do not need to have its guise. On WhatsApp, such materials often imitate firsthand accounts or testimonies – appearing as informal yet persuasive reports from individuals who claim to have witnessed or experienced a specific event; thus, there is a wide range of texts, audios and videos, especially in the first person (Demuru et al., 2021; Fernandes et al., 2019).

The inherently private nature of WhatsApp fosters a more personal mode of communication, particularly effective when messages come from trusted individuals such as family members or close acquaintances. In contrast, Facebook content tends to be more public, even when shared within limited privacy settings, offering broader visibility and engagement. It is plausible, though it requires further investigation, that content shared on Facebook often resembles traditional news structures, incorporating features such as leads, inverted pyramid format, cited sources, multimedia elements, and hyperlinks; in other words, it is content that imitates journalistic presentation. This resemblance also applies to the websites hosting such content, which often adopt a layout and tone suggestive of legitimate news outlets. This creates a paradox: despite journalism’s declining credibility in certain circles, its structural conventions are still borrowed to lend legitimacy to disinformation.

Even when disinformation lacks a formal news structure, one cannot deny that “the dissemination of false information disguised as news affects the public’s perception of journalistic credibility” (Santos & Maurer, 2020, p. 2); thus, disinformation both contributes to and capitalizes on journalism’s credibility crisis. Both fake news and fact-checking agencies leverage the remaining credibility of journalism, albeit diminished for some audiences, whether by mimicking the structural conventions of journalism in the case of disinformation or by applying verification methodologies in the case of fact-checking. This dynamic is also a result of the transformation of journalism itself, which has lost its once-exclusive authority over the dissemination of public information in the digital era, leading to content that is increasingly declarative and dependent on source statements, often shaped by the pressures of real-time publication.

News organizations, thus influencing the broader journalistic institution, seem to weaken their authority when they delegate post-publication verification to external agencies, becoming increasingly dependent on source narratives, regardless of their accuracy. Journalism appears caught between the rapid dissemination afforded by social media and the more methodical verification processes of fact-checking agencies, without providing either the immediacy of the former or the analytical rigor of the latter.

Guerra (2005) offers a useful distinction between the journalistic institution and the journalistic organization, as they refer to different dimensions of journalism, highlighting that criticisms directed at one do not necessarily apply to the other. The institution encompasses the normative principles that define journalism’s ideals of journalism, while the organization refers to the operational rules that guide the execution of journalistic practices.

The institution embodies the fundamental principles of journalism and its social function. The organization, in turn, serves as the concrete manifestation of those principles—applying institutional norms in practical and real-world contexts. The organization refers to the people and means that it puts into practice its concept, ideology and function imbued in the institutionalization of journalism. What is considered social recognition as a function of journalism is not necessarily the same as what is actually produced in organization; in other words, “the first belongs to the normative dimension, while the second belongs to an empirical nature” (Seixas, 2009, p. 73). Dedicating to journalistic institutions means understanding the proliferation of large (media conglomerates) and small organizations, realizing that each organization is free to absorb the rules and norms in its own way.

However, there may be a decrease in the circulation (or effects) of fake news may be occurring in Brazil, mainly due to a partnership of technologies (especially social media) with fact-checking agencies, as the then President of the Superior Electoral Court (TSE), Minister Luís Roberto Barroso, said during a seminar on the elections: “We think that this is an election in which the level of fraudulent news circulation was minimal”. In the 2018 elections, disinformation served as an instrument of mobilization, especially because digital media gained undeniable relevance in the decision-making process, highlighting mainly that the winning candidate had only eight seconds in the Free Electoral Broadcast Airtime and did not attend TV debates after being stabbed.

Since 2016 in the USA, Facebook has implemented a type of algorithmic penalty by reducing the reach of suspicious content, reducing, according to Moreno & Moutinho (2018), 80% of the distribution of material considered harmful by the company’s fact-checking partner agencies. The programme arrived in Brazil in 2018, especially for that year’s elections. The company uses nine gradations of labels, of which, ‘false’, ‘mixture’ and ‘false headline’ are the most serious and cause a reduction in propagation, prohibition of paid promotion and notification, and may result in penalties in case of recurrence.

Account administrators may request a review of the content checked by the agencies, as occurred in cases of members of the Free Brazilian Movement (MBL), who generally publish unreliable information and accuse the agencies of being subordinate to political ideologies, even though Lupa and Aos Fatos, Facebook partners, are part of the IFCN, thus, openly non-partisan. In Brazil, these partnerships resulted in the rejection of half of these mentions on social media (Moreno & Moutinho, 2018), although, at Lupa, there is a balance in the verification of left and right-wing politicians (Santos, 2019).

5. Conclusion

Fake news often circulate widely and play a significant role in the contemporary communication flow, especially on the internet and, more specifically, on social media platforms. In this disinformation process, it is crucial to identify the patterns of fake news, particularly when they target vulnerable groups such as indigenous peoples. Therefore, understanding not only the frequency but also the topics, origins, and formats of disinformation highlighted by the fact-checking agency Lupa is essential. Regarding the origin of fake news, social media has proven to be a priority area (89%, 32) for its dissemination, especially Facebook (72%, 26) and, since the COVID-19 pandemic, also WhatsApp. Fake news circulation on social media has grown to such an extent that these platforms have had to adopt a role of fact-checking, an approach that Lupa has strengthened since 2018, notably due to its partnership with Meta.

However, fact-checking agencies are not the only entities focusing on disinformation within social media; the platforms themselves have recognized the issue of fake information spreading freely on their networks. As a result, they not only label some content as misleading, such as on Facebook and Instagram, but also limit the sharing of potentially fake information through notifications, removals, or demonetization. Moreover, these platforms have partnered with fact-checking organizations; for example, Lupa began providing fact-checking services to Facebook. These efforts help reduce the circulation of fake news, at least on social media, which, according to our data analysis, remains the main space for its dissemination.

Blogs and news websites, on the other hand, exhibit a lower density of information disorder, especially when professional journalists conduct thorough verification and investigation. The spread of disinformation is more effective in outlets that rely on so-called declarative journalism, which is based solely on the statements of interviewees without verification or fact-checking, thereby undermining institutional credibility (Dourado, 2019; Martins & Teixeira, 2025; Santos & Maurer, 2020). Nevertheless, the press has been evaluated by Lupa on three occasions. Unlike the disinformation that circulated widely on social media, these investigations were initiated by Lupa’s professionals rather than by public demand. The first two, conducted in 2018 and 2019, analysed speeches by former President Jair Bolsonaro presented in text format; the third, in 2023, involved a full video interview with the President of the Chamber of Deputies on the television programme Roda Viva, in which falsehoods were identified.

Although journalism as an institution has faced credibility challenges that facilitate the spread of disinformation, some of its supposed articles appear in text format, alone or accompanied by images, circulating on social media, especially Facebook. This platform, in fact, has a direct connection with textual disinformation (25 out of 26 cases), serving as a primary format to attract attention in cases of falsehoods disseminated about indigenous peoples. Thus, text plays a leading role not only in spreading information disorder but also in its (de)contextualization. On WhatsApp, however, disinformation predominantly circulates through video, present in all six cases analyzed, regardless of accompanying text.

Lupa is an independent and one of the leading fact-checking agencies in Brazil; however, the data presented may reflect its editorial stance, particularly regarding the emphasis it places on disinformation against Brazilian indigenous peoples. It is possible that other agencies, especially those affiliated with traditional media where indigenous issues receive less coverage, might report different figures, potentially lower. Nevertheless, the data in this study demonstrate that social media demands attention not only from fact-checking agencies but also from all users who share information on these platforms, especially in text and video formats. These findings could also support further analysis of the components of media framing, such as interpretation, evaluation, and recommendation, and suggest the usefulness of combining content analysis with framing methodology.

6. Funding

This work was supported by the CNPq – National Council for Scientific and Technological Development under Grant 405660/2023-5.

7. Statement on the use of artificial intelligence

Artificial intelligence has not been used in this article.

Biography of the author

Allysson Viana Martins is professor of the Postgraduate Programs in Communication, PPGCOM, and in Regional Development and Environment, PGDRA, and coordinator of Digital Media and Internet Laboratory, MíDI, at Federal University of Rondônia, UNIR, Brazil. Postdoctorate in Communication at Federal University of Ceará. PhD in Communication and Contemporary Culture at Federal University of Bahia. He is the author of the books in portuguese “It Wasn’t Just a Little Flu, But Disinformation” (2025), “Journalism and Battles of Memories in the 50th Anniversary of the 1964 Coup” (2020) and “Digital Journalism between Memory Networks on the 9/11 Commemoration” (2022).

References

Borges, B., Gomes, L., & Maia, F. (2022, October 29). O combate às fake news nas aldeias. Jota. https://encurtador.com.br/8X5ec

Carvalho, C., López, M., & Andrade, K. (2019). Agências de checagem no Brasil: uma análise das metodologias de Fact-Checking. Anais do 17º Encontro Nacional de Pesquisadores em Jornalismo. UFG, Goiânia. https://encurtador.com.br/5lgIK

Cazetta, J., & Reis, A. (2019). As fontes dos serviços de fact-checking luso-brasileiros. Comunicação Pública, 14(27), 1-18. https://doi.org/10.4000/cp.5348

Cruz, M., & Barbosa, B. (2023, June). “Tiveram medo da vacina”. Cultura indígena bate de frente com orientações de combate à covid e notícias falsas sobre imunizantes. UOL. https://encurtador.com.br/n9u8R

Demuru, P., Fechine, Y., & Rodrigues, C. (2021). Desinformação como camuflagem: modos de produção da verdade no WhatsApp durante a pandemia. Anais do 30º Encontro Anual da Compós, São Paulo: PUC-SP. https://encurtador.com.br/cyiAo

Derakhshan, H., & Wardle, C. (2017). Information disorder: definitions. In: Understanding and Addressing the Disinformation Ecosystem. Pensilvânia: Annenberg School for Communication (pp. 5-12). https://encurtador.com.br/C14GF

Dourado, T. (2019). Fact-checking como possibilidade de media accountability sobre o discurso político? Uma análise de conteúdo das iniciativas Aos Fatos, Lupa e Truco. Revista Compolítica, 9(2), 93-112. https://encurtador.com.br/poOfa

Dourado, B. (2024. April 24). Ranking: as redes sociais mais usadas no Brasil e no mundo em 2023, com insights, ferramentas e materiais. RD Station. https://encurtador.com.br/Dzhl6

Eichler, V., Kalsing, J., Gruszynski, A. (2018). O ethos do jornal O Globo e a campanha contra as fake news. Media & Jornalismo, 32(18), 139-154. https://encurtador.com.br/V49nU

Entman, R. (1993) Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43(4), 51-58. https://doi.org/10.1111/j.1460-2466.1993.tb01304.x

Fernandes, C., Oliveira, L., & Gomes, V. (2019). Tensionamentos entre campos sociais: as fake news e a reconfiguração do campo comunicacional e político na era da pós-verdade. XXVIII Encontro Anual da Compós. PUC-RS, Porto Alegre. https://encurtador.com.br/BlV6r

Gehlen, M. (2018). Fact-checking: o caso da Lupa, a primeira agência de checagem de notícias do Brasil. Estudos de Jornalismo, (9), 44-60. https://www.revistaej.sopcom.pt/ficheiros/20190103-ej9_2018.pdf

Graves, L., Cherubini, F. (2016). The Rise of Fact-Checking Sites in Europe. Reuters Digital News Project. https://shre.ink/t6je

Gruszynski, A., Kalsing, J., Hoewell, G., & Brandão, C. (2020). Fact-checking e saúde: análise da seção ‘Verdade ou Boato’ de GaúchaZH. Revista Eletrônica de Comunicação, Informação & Informação em Saúde, 14(1), 51-71. https://doi.org/10.29397/reciis.v14i1.1860

Guerra, J. (2005). Instituição e organização jornalística: uma distinção conceitual. XXVIII Congresso Brasileiro de Ciências da Comunicação (Intercom), UERJ, Rio de Janeiro. https://shre.ink/t6jk

LUPA fecha parceria com Facebook e vai checar posts feitos na plataforma (2018, May 18). Lupa. https://shre.ink/t6jU

LUPA expande comunicação com leitores com canal no WhatsApp (2020, October 22). Lupa. https://encurtador.com.br/b5kUL

Macário, C. (2023, November 17). ‘Índio de iphone’ e ‘borracha nessa corja’: discurso de ódio e fakes escancaram violência contra indígenas em SC. Lupa. https://shre.ink/t6GK

Madeiro, C. (2023, October 20). MPF pede afastamento de prefeito por fake news contra ação em área indígena. UOL. https://shre.ink/t6G3

Mansani, T. (2022, February 08). Jovens jornalistas indígenas combatem fake news na Amazônia. DW Brasil, publicado em 08 fev. https://p.dw.com/p/46gUq

Marés, C. (2018, December 06). É verdade que Bolsonaro elogiou cavalaria norte-americana por dizimar índios. Lupa. https://shre.ink/t6hT

Marko, K., Neves, P., Reinholz, F. (2022, October 11). Áudios com fake news espalham medo sobre terras indígenas e quilombolas no interior gaúcho. Brasil de Fato. https://shre.ink/t6uG

Martins, A. (2024). Politização das mentiras sobre a população indígena na agência digital de checagem Lupa. Comunicação & Inovação, São Caetano do Sul, São Paulo, (25), 1-20. https://doi.org/10.13037/ci.vol25.e20249757

Martins, A. (2025). Padrão da desinformação contra os povos indígenas: enquadrando as fake news checadas pela agência Lupa. 34º Encontro Anual da Compós, Curitiba. https://shre.ink/t6uw

Martins, A., & Teixeira, J. (2023) Checagens sobre a covid-19 e enquadramento temático nas agências Fato ou Fake e Lupa. Contracampo, (42), 1-16. https://doi.org/10.22409/contracampo.v42i3.59028

Martins, A., & Teixeira, J. (2025). Não era só uma gripezinha, mas desinformação. Florianópolis: Insular. https://shre.ink/t6hU

Matthes, J., & Kohring, M. (2008). The Content Analysis of Media Frames: Toward Improving Reliability and Validity. Journal of Communication, (58), 258-279. https://doi.org/10.1111/j.1460-2466.2008.00384.x

Meneses, J. (2018). Sobre a necessidade de conceptualizar o fenómeno das fake news. Observatório (OBS*). 12(5), 37-53. https://shre.ink/t6hB

MJSP desmente fake news sobre indígenas Yanomami. (2023, January 27). Gov.br, Ministério da Justiça e Segurança Pública. https://shre.ink/t6hg

Moreira, R. (2023, January 25). Fala de deputado bolsonarista de MS sobre indígenas Yanomami repercute nas redes sociais. G1. https://shre.ink/t6ht

Moreno, T., & Moutinho, N. (2018). Facebook e fact-checkers: o caso da Agência Lupa. Ameaças ao ciberjornalismo. Atas do VI Congresso Internacional de Ciberjornalismo. Porto, Portugal, 92-105. https://hdl.handle.net/10216/120956

Rômany, Í. (2023, September 22). É falso que eleição foi fraude e que Alexandre de Moraes declarou guerra à população. Lupa. https://shre.ink/t6h3

Santos, C., & Maurer, C. (2020). Potencialidades fact-checking no combate à desinformação. Comunicação & Informação, Goiânia, (23), 1-14. https://doi.org/10.5216/ci.v23i.57839

Santos, L. (2019). Fake news e fact-checking: as movimentações no campo comunicacional vistas a partir da midiatização da sociedade e da objetividade jornalística. Anais do 17º Encontro Nacional de Pesquisadores em Jornalismo (SBPJOR). UFG, Goiânia. https://shre.ink/t6hK

Schudson, M.; & Zelizer, B. (2017). Fake News in Context. Understanding and Addressing the Disinformation Ecosystem. Pensilvânia: Annenberg School for Communication (pp. 1-4). https://shre.ink/t6hX

Seixas, L. (2009). Por uma outra classificação: gêneros discursivos jornalísticos e gêneros discursivos jornálicos. Revista Galáxia, 9(18), 70-84. https://shre.ink/t6hN

Silva, M., & Melo, S. (2020). Fake news: fronteiras do jornalismo e circulação de (des)informação sobre saúde. Revista Eletrônica de Comunicação, Informação & Informação em Saúde, 14(1), 1-5. https://doi.org/10.29397/reciis.v14i1.2047

Silva, M., Albuquerque, M., & Veloso, M. (2019) Representação da informação noticiosa pelas agências de fact-checking: do acesso à informação ao excesso de desinformação. Revista Brasileira de Biblioteconomia e Documentação, 15(2), 410-426. https://shre.ink/t6h4

Soares, M. (2009). Análise de enquadramento. In: Barros, A., & Duarte, J (Eds.). Métodos e técnicas de pesquisa em Comunicação (pp. 450-465). Atlas.

Spinelli, E., & Santos, J. (2018). Jornalismo na era da pós-verdade- fact-checking como ferramenta de combate às fake news. Revista Observatório, 4(3), 759-782. https://doi.org/10.20873/uft.2447-4266.2018v4n3p759

Tandoc Junior, E., Lim, Z., & Ling, R. (2018). Defining “Fake News”: a typology of scholarly definitions. Digital Journalism, 6(2), 137-153. https://doi.org/10.1080/21670811.2017.1360143

Teixeira, J., & Martins, A. (2020). Fact-checking no combate às fake news sobre a COVID-19: um estudo exploratório das agências digitais de checagem de fatos contra a desinformação da pandemia. Comunicação & Inovação, (21), 63-81. https://shre.ink/t6GH

Teixeira, J., & Martins, A. (2022). Thematic patterns of disinformation about COVID-19: the framing of checks in the Fato ou Fake and Lupa agencies. Journalism and Media, (3), 27-39. https://doi.org/10.3390/journalmedia3010003

Valentim, F. (2019). Fact-checking como possível ferramenta qualificadora do debate público. Caderno da Escola Superior de Gestão Pública, Política, Jurídica e Segurança, 2(1), 197-215. https://shre.ink/t6GB

Vimieiro, A., & Dantas, M. (2009). Entre o explícito e o implícito: proposta para a análise de enquadramentos da mídia. Lumina, 3(2), 1-16. https://periodicos.ufjf.br/index.php/lumina/article/view/21048

Vimieiro, A., & MAIA, R. (2011) Análise indireta de enquadramentos da mídia: uma alternativa metodológica para a identificação de frames culturais. Famecos, 18(1), 235-252. https://doi.org/10.15448/1980-3729.2011.1.8810

Wang, C. (2020). Fake News and related concepts: definitions and recent research development. Contemporary Management Research, 16(3), 145-174. https://doi.org/10.7903/cmr.20677