Bolivia: Decoding the Patterns of Electoral Disinformation

The 2025 general elections in Bolivia confirmed that disinformation is no longer a marginal phenomenon but has become a central player in public debate. Under the shadow of polarization and the memory of the 2019 conflict, the deployment of false narratives, manipulated content, and coordinated attacks was present from the beginning of the election campaign until the second round of voting. ProBox and Bolivia Verifica combined their methodologies and analytical capabilities to assess the scale of this phenomenon.

With ProBox’s analysis of around 12,000 social media posts and Bolivia Verifica’s systematic monitoring of more than 200 pieces of content identified between May 19 (the start of the election campaign) and October 19 (the day of the second round), it was possible to observe disinformation in terms of its volume and scope, as well as its narratives, resources, and political objectives, which ranged from attempting to influence voting intentions and spread narratives of fraud to discrediting political actors, attacking the media, and attempting to erode confidence in the “Tribunal Supremo Electoral-TSE” (Supreme Electoral Tribunal).

Far from being isolated episodes, this content followed clear patterns of adaptation that escalated strategically at the most critical moments of the contest. Understanding these dynamics is key to anticipating risks and strengthening democratic resilience in a scenario where information manipulation continues to evolve.

Negative narratives increase with the electoral situation

Between May 19 and October 19, 2025, Bolivia Verifica identified 254 pieces of disinformation divided into six main themes, the three main ones being: 

1. Misleading propaganda: this was the main theme, accounting for 59% of the false content detected; the main objective was to confuse the electorate and modify their perception of the candidates in order to influence the vote, mainly through the use of false statements in graphic or audiovisual format, especially about those who were leading in the official polls.

2. Data manipulation: this was the second most recurrent theme, accounting for 15% of all content identified. This strategy also focused on confusing the electorate through the deliberate modification of real information, especially through the use of statements taken out of context and official poll data with erroneous explanations.

3. Personal attacks: this topic ranks third with 14% of the content identified, mainly focused on defaming the origin, ideology, or private life of candidates to disqualify them or generate mass rejection. 

The rest of the topics registered much lower percentages: narratives of fraud (6%), hate speech (4%), and direct attacks on the government (2%).

The ProBox study focused on the period between August 1 and November 30, 2025, analyzing 10,422 mentions on X and 2,410 posts on TikTok related to the electoral process and the topics identified by Bolivia Verifica. It was found that 88.9% of the conversation was dominated by negative narratives focused on hate speech, regionalism, and polarization (47.5%); In all mentions, the main protagonists identified were President-elect Rodrigo Paz Pereira with 24%, followed by former presidential candidate Jorge “Tuto” Quiroga Ramirez with 22%, and Vice President-elect Edmand Lara Montaño with 18%.

At the same time, an alarming relevance was identified in narratives of fraud and attacks on electoral institutions, accounting for 34.7% of the content analyzed. This conversation was heavily influenced by historical resentments linked to the 2019 elections, attacks on the media, and persistent discourse of mistrust toward the TSE.

Both the findings of Bolivia Verifica and the analysis by ProBox show similar patterns in online conversation, especially in negative narratives, personalized attacks, and questioning of the electoral process. 

Another finding shared by both organizations is the reactive nature of disinformation: content increased in line with milestones in the electoral calendar, proclamations, and news events. 

In the days leading up to the first round, attacks focused on the frontrunner in the polls; after the August 17 vote, the narrative quickly shifted to Rodrigo Paz Pereira, the winner of that round, and to direct questioning of the TSE. The same pattern was repeated in October during the second round, showing that disinformation evolved during the process and revolved around the immediate political event.

Peaks of activity recorded:

• August 17 (first round): Bolivia Verifica recorded one of the highest volumes of debunkings of the process, with nearly 10 pieces of false content identified that day. At the same time, ProBox counted 380 mentions focused on allegations of fraud and false polls.
• August 18: ProBox recorded the highest peak of digital conversation, with 516 mentions dominated by accusations of fraud and attacks on the TSE.
• August 19: Conversation remained high and polarized.
• October 19 (second round): Bolivia Verifica again reported around 10 debunked claims, making that day one of the highest in terms of the circulation of hoaxes.
• October 20: ProBox identified 303 mentions, again driven by allegations of fraud and institutional questioning.

Overall, the data shows a clear pattern: the days of voting and the days immediately following saw both the highest volume of conflictive digital conversation and the highest number of false content detected and debunked. In other words, the moments of electoral decision acted as simultaneous triggers for disinformation and polarization.

Statements, fake polls, and smear campaigns: the most commonly used tactics

Bolivia Verifica’s analysis of disinformation content throughout the electoral process revealed that the most commonly used element was false statements attributed to candidates or other social actors. This was used strategically with the aim of subtracting or adding votes to a particular candidate. 

The second most commonly used tactic was fake polls, with a higher volume of these identified in the first round (35 fake studies), while in the second round, the number dropped to 9. This method encompasses various forms of deception: from fabricated polls (created entirely from scratch, with unsubstantiated data) and manipulated polls (alteration of the results of a real poll published by its original authors) to the use of unregulated polls (surveys shared on social media without verifiable reliability, methodology, or legality; prohibited by the country’s electoral regulations).

In terms of formats, graphic content (images and montages) was the most widely used by disinformers. This tactic consisted of using photographs taken out of context or designing images that imitated the graphic style of real media outlets—both national and international—in order to lend credibility to the falsehood. Red Uno was the media outlet whose graphic style was replicated most frequently, followed by Unitel and El Deber. It is also important to note that the use of artificial intelligence to generate this false content was among the highest in the region, a phenomenon discussed in detail in the Bolivia Verifica special report.

Likewise, the distribution of this content was carried out especially on TikTok; disinformers designed a specific tactic for this platform: they uploaded the fake image along with a musical base to convert it into a valid audiovisual format that, in essence, remained a static image. This adapted content was easily amplified on TikTok and then replicated on other networks such as Facebook and WhatsApp, ensuring maximum dissemination.

Within the sample analyzed by ProBox, a persistent narrative of mistrust toward the press was consolidated, operating both through open attacks and the misuse of journalistic brands to lend credibility to false content.

Unitel (91 mentions), El Deber (64 mentions), and Red Uno (11 mentions) were also identified in the digital monitoring as the most cited media outlets, especially with narratives about being “sold out” or “manipulating polls.” Unitel was the most exploited in the narrative of circulating false polls. El Deber appeared less linked to impersonated content, but it was incorporated into discourses that questioned its editorial independence. 

Que pasó con el "boca de urna" de Captura Consulting?
Hubo la órden de esconderlo?
Seguramente ganaba Tuto, y prefirieron evitar la fatiga.
El Deber y Red Uno tienen una deuda con el país.

— 🅷🅴🅲🆃🅾🆁 🅲🅰🆂🆃🆁🅾 🅶. (@HequeCastro) October 20, 2025

The conversation on TikTok was marked by an emotional and fragmented tone, dominated by narratives of suspicion, accusation, and delegitimization. Although the more than 2,000 videos analyzed contained informative or neutral content with a wide reach, the greatest visibility was concentrated on pieces that questioned the credibility of the electoral process, political actors, and institutions, as well as the dissemination of unsubstantiated polls and allegations. TikTok thus functioned as a space for dispute and polarization, where videos appealed to denounce, warn, or ridicule, amplifying feelings of mistrust and resistance to the electoral process.

@cris.nation FRAUDE ELECTORAL EN BOLIVIA🇧🇴 La Dictadura del MAS en Bolivia y sus aliados Rodrigo Paz y Edman Lara (FALSA OPOSICIÓN) hicieron fraude en Bolivia, hubo muchas denuncias de irregularidades en los recintos electorales, y en menos de 3 horas de la recepción de votos se dió por ganador a PAZ-LARA siendo que en el cómputo de la OEP estaba ganando Tuto Quiroga y una hora antes de declarar la supuesta "Victoria" de los aliados de Evo Morales, se cortó la luz y se paralizó el cómputo de votos y el porcentaje de votos se invirtió a favor de los socialistas PAZ-LARA con un 54.5% de votos y Tuto Quiroga con un 45.5%. El programa de televisión UNITEL fue cómplice de este circo y show masista dando por ganador al Binomio sin tener los resultados finales. JAVIER MILEI, DONALD TRUMP, PEÑA Y OTROS LÍDERES DE DERECHA FUERON ENGAÑADOS Y PENSARON QUE LA DERECHA GANÓ EN BOLIVIA, SIENDO TODO LO CONTRARIO. Una izquierda comunista radical que el pueblo boliviano en su gran mayoría rechaza y repudia. ¡DIOS TENGA MISERICORDIA DE BOLIVIA!🇧🇴 ¡DIOS SALVE BOLIVIA! @realdonaldtrump @laderechadiario ♬ sonido original – MILEI ON TOP

At least 84 videos were detected on this platform promoting theories of media manipulation, repeatedly pointing to Unitel, Red Uno, and El Deber as accomplices in false polls, covert propaganda, or artificially inflating candidacies. These narratives were repeated in multiple audiovisual pieces, reinforcing the idea that no measurements or coverage were reliable.

@sowig25

Otra vez UNITEL y RED UNO inventando ganadores… ¡Ya cansan ! Por eso votaré por Manfred Yo no voto por falsas encuestas de Red Uno y UNITEL Voto castigo para Samuel y Tuto Manfred Presidente!!! Samuel y Tuto no son mi opción, por más que UNITEL mienta que están primeros SAMUEL Y TUTO PAGAN ENCUESTAS Y NO ME DEJARE LLEVAR POR MENTIRAS… “Que RED UNO y UNITEL se guarden sus números de la mentira, el pueblo piensa distinto…Manfred Presidente!!! UNITEL y RED UNO nos quieren meter a Samuel y Tuto por los ojos… ¡pero esoo no va a pasar! Las encuestas de UNITEL y RED UNO parecen escritas en la oficina de algún candidato Basta de manipulación UNITEL y RED UNO. Nooo ME TRAGO ESE CUENTO de Tuto y Samuel están primeros !!! No necesito que Red Uno y UNITEL me digan por quién votar. No Samuel ni Tuto… Manfred Presidente!!! Cada vez que UNITEL y RED UNO sacan su encuesta, pierden más credibilidad…por eso no votaré por ellos. UNITEL: la máquina de inflar candidatos reciclados como Samuel y Tuto El pueblo no es t0nt0 Samuel y Tuto no nos representan con falsas encuestas … Las encuestas de UNITEL son como un guion viejo: siempre con los mismos protagonistas… 14. RED UNO y UNITEL… siempre la misma mentira…nadie cree esas encuestas !! UNITEL y RED UNO hacen encuestas para convencer, no para informar … Por eso mi voto será por Manfred !! Yo decido en la urna, no en pantallas de RED UNO y UNITEL NO A LAS FALSAS NOVELAS DE ENCUETAS DE RED UNO Y UNITEL UNITEL y RED UNO MANIPULAN ENCUESTAS por eso mi voto será por el Bom Bom Encuestas de RED UNO y UNITEL = propaganda disfrazada que ya nadie cree ENCUESTAS DE RED UNO Y UNITEL = FALSEDAD QUE LOS BOLIVIANOS QUE NO CREEMOS UNITEL podrá inventar números, pero no podrá cambiar mi voto…MANFRED PRESIDENTE C4R4J0000!!

♬ sonido original – 𝗩_𝗠𝗨𝗦𝗜𝗖𝗔

Expressions such as “sold-out press” (62 mentions) and “bought media” (22 mentions) were also recorded; although these publications did not reach a significant scale, they contribute to interests that seek to reinforce a climate of generalized suspicion about the role of the media and journalists.

Weaponizing Fraud: How the Narrative Against the TSE Was Built

The allegation of “electoral fraud” was one of the most persistent themes in the digital conversation analyzed by Probox, with direct attacks on the “Tribunal Supremo Electoral-TSE” (Supreme Electoral Tribunal). from the first round and a peak during the second round. The narrative was based on unresolved resentments from 2019 and a structural distrust of institutions, reinforcing messages such as “the TSE decides who governs,” “the elections are rigged,” or that there is “MAS fraud.”

Some key data from the monitoring:

• 3,612 mentions are classified as fraud and institutional attacks.
• The term “electoral fraud” appeared in 782 mentions, many of which reactivated comparisons with previous processes.
• The “null vote” accounted for at least 2,850 mentions in the second round, becoming the main narrative node.
• Nearly 60 pieces of content directly promoted the idea of fraud, some with an impact outside of social media by motivating mobilizations.
• The most repeated hashtags on the topic were #Fraude, #FraudeElectoral, and #FraudeBolivia.
• More than 50% of the content monitored by Bolivia Verifica between the day of the second round and the following two weeks was directed against the credibility of the TSE.

On the one hand, Evo Morales’ promotion of null votes after being eliminated from the race intensified digital polarization. His calls to “mark the ballot” were interpreted by his supporters as a legitimate form of protest and by his detractors as a strategy to sabotage the democratic process.

El gobierno de Arce destruyó la economía dividió y secuestró nuestro instrumento político con sentencias judiciales. Proscribió a las siglas que apoyaron nuestra candidatura y ahora, bajo un cerco mediático y político, pretende cuestionar a la gente que decide votar nulo.

Todo… pic.twitter.com/DJtNGtLNdX

— Evo Morales Ayma (@evoespueblo) July 31, 2025

There were also repeated accusations that “Masismo” was responsible for possible irregularities, while figures such as Samuel Doria Medina were also accused of being part of alleged fraudulent strategies. Morales remained one of the main sources of polarization and mutual accusations.

El Fraude electoral masista es la vieja práctica de imponer gobernantes desde el TSE 😎 https://t.co/p113tQ3EJW

— Lisberth Vargas (@Lisprensalibre) October 23, 2025

On TikTok, the narrative took on a highly emotional and visual tone. Videos circulated denouncing alleged “systematic electoral fraud,” including theories about strategic blackouts during the count to alter results and favor alliances linked to the ruling party to the detriment of candidates such as Tuto Quiroga.

@gocommunicationbo ⚠️ Jóvenes de la Generación Z en La Paz manifiestan que no permitirán, lo que ellos llaman, un "fraude electoral"🚫 #generaciónz #eleccionesbolivia #noticiasbolivia #bolivia #tsebolivia ♬ sonido original – Go Communication Bolivia

 

Around 44 videos sought to reinforce widespread mistrust of the justice system and institutions, associating them with corruption, manipulated polls, and alleged “state capture.” Street polls presented as the authentic voice of the “people” amplified this climate of polarization. There were also pieces that directly questioned the credibility of the TSE, casting doubt on the transparency of the voter registry, the electoral software, and the audit processes, as well as content showing demonstrations and protests in front of electoral courts, with the intention of creating the perception of a sustained institutional conflict.

@cris.nation FRAUDE ELECTORAL EN BOLIVIA🇧🇴 La Dictadura del MAS en Bolivia y sus aliados Rodrigo Paz y Edman Lara (FALSA OPOSICIÓN) hicieron fraude en Bolivia, hubo muchas denuncias de irregularidades en los recintos electorales, y en menos de 3 horas de la recepción de votos se dió por ganador a PAZ-LARA siendo que en el cómputo de la OEP estaba ganando Tuto Quiroga y una hora antes de declarar la supuesta "Victoria" de los aliados de Evo Morales, se cortó la luz y se paralizó el cómputo de votos y el porcentaje de votos se invirtió a favor de los socialistas PAZ-LARA con un 54.5% de votos y Tuto Quiroga con un 45.5%. El programa de televisión UNITEL fue cómplice de este circo y show masista dando por ganador al Binomio sin tener los resultados finales. JAVIER MILEI, DONALD TRUMP, PEÑA Y OTROS LÍDERES DE DERECHA FUERON ENGAÑADOS Y PENSARON QUE LA DERECHA GANÓ EN BOLIVIA, SIENDO TODO LO CONTRARIO. Una izquierda comunista radical que el pueblo boliviano en su gran mayoría rechaza y repudia. ¡DIOS TENGA MISERICORDIA DE BOLIVIA!🇧🇴 ¡DIOS SALVE BOLIVIA! @realdonaldtrump @laderechadiario ♬ sonido original – MILEI ON TOP

This pattern was also confirmed by Bolivia Verifica: more than 50% of the disinformation content monitored between the day of the second round and the following two weeks was specifically directed against the credibility of the TSE, identifying nearly 60 pieces of content that directly promoted the idea of electoral fraud; several of them with narratives calling for the results to be disregarded or warning of alleged ongoing manipulation. 

Both analyses agree that the narrative of fraud not only remained constant throughout the campaign, but intensified in the second round and in the post-election period, consolidating itself as one of the predominant narrative trends and further eroding institutional trust.

Who generated the wave of disinformation? 

The wave of disinformation that marked the electoral period was neither spontaneous nor homogeneous. Cross-referencing the findings of Bolivia Verifica and ProBox reveals a more complex ecosystem: identifiable actors who issued or amplified false content, and a digital conversation that amplified it until it became the dominant frame of interpretation. 

Bolivia Verifica identified at least 171 actors in the 254 pieces of content, classifying them into two main groups: the first consists of recognized accounts or individuals who publicly issued or disseminated false information; the second group includes profiles with no real credibility, anonymous or without a track record, whose sole function appears to be to create and massively amplify disinformation content.

26 actors belong to the first group that issued or amplified false content that was debunked by Bolivia Verifica. These actors are mainly politicians and media outlets, notably current Vice President Edmand Lara and Radio Kawsachun Coca, the latter known for disseminating content in favor of socialism or former President Evo Morales. More details on this can be found here.

145 profiles belong to the second group and published at least one piece of false content during the election period, mainly on TikTok. Fourteen of these 145 profiles were repeat offenders, publishing disinformation two or more times. 

But how did this content manage to transcend its original accounts and become a mass conversation? ProBox’s analysis shows that disinformation found fertile ground in a network of individual accounts that replicated, adapted, and reinforced the same narrative frameworks. It was not just a question of who published first, but who managed to amplify it best.

The conversation was driven by individual accounts, which were responsible for 99.74% of the more than 12,000 mentions analyzed. In this ecosystem, some accounts function as amplification nodes, concentrating visibility, interactions, and narrative traction.

These accounts did not act identically or respond to a single ideological line, but they shared a common trait: their ability to establish interpretive frameworks—fraud, illegitimacy, betrayal, regional confrontation—that were then replicated by other users.

Together, they did not form an automated network or a clearly identifiable centralized operation. However, their activity shows patterns of semi-coordinated amplification, in which similar messages, shared narrative frameworks, and coinciding publication times contributed to the escalation of certain narratives. 

The effect was not to impose an alternative truth, but to saturate the digital public space with mistrust, anger, and suspicion, hindering the circulation of verified information. 

The fight against disinformation needs reinforcements

The Bolivian election left a clear lesson: disinformation adapts quickly and occupies key spaces in public debate, even in the face of institutional efforts to contain it.

The “Órgano Electoral Plurinacional-OEP” (Plurinational Electoral Body) disseminated graphic and audiovisual materials to refute false content during the campaign (examples: 1, 2, 3, 4); but even so, narratives of fraud and institutional attacks managed to take hold in online conversation. At the same time, the United Nations Development Programme (UNDP) promoted the creation of the “Coalición Nacional contra la Desinformación Electoral” (National Coalition against Electoral Disinformation), an effort that brought together civil society, the media, international partners, and public entities such as the OEP to strengthen institutions and mitigate hate speech; Bolivia Verifica was part of this alliance. In turn, various media outlets also warned about the circulation of false content, relying on data from Bolivia Verifica, the Coalition, and electoral experts (examples: 1, 2, 3, 4, 5).

Even ahead of the second round, candidates Rodrigo Paz Pereira and Jorge Quiroga signed a public commitment to a clean campaign free of disinformation. 

Despite these actions, the findings of ProBox and Bolivia Verifica reveal pending challenges. Platforms such as Telegram and X were partially left off the radar due to technical and resource limitations, and the inability to directly tag or report false content within Meta and TikTok’s own networks reduces mitigation capacity.

The process also showed that tackling disinformation is not only a technological challenge, but also a social and educational one. Rural populations and older adults require greater support, while other sectors need to strengthen critical skills in the face of emotional narratives, media impersonation, and content powered by artificial intelligence. The combination of verification, media literacy, and institutional cooperation is emerging as a key condition for protecting public debate in future electoral processes.

This report was produced by monitoring and analyzing 10,422 mentions on Platform X (formerly Twitter), 2,410 videos collected on TikTok, as well as a qualitative analysis of content circulated on Facebook by ProBox, and the compilation of content monitoring carried out by Bolivia Verifica in the last presidential elections.

Go to all articles