Deepfakes, Identity Politics, and Narrative Attack Campaigns Target the Mexican Election

In a historic first, Mexico just elected Claudia Sheinbaum, its first female and first Jewish president, prevailing over a wave of narrative attacks.

Women now hold significant positions of power in the government and comprise half of Congress and the Cabinet. Voters turned out in large numbers to vote for officials in nearly 20,000 government positions. In addition to electing a new president, 500 Chamber of Deputies members, 128 senators, and nine governors, all 32 Mexican states held concurrent elections for thousands of local seats—around six times more posts than during the previous election in 2018. Of Mexico’s population of around 130 million, roughly 99 million were eligible to vote, with young people comprising the largest voting bloc at 26.5%. 

Deepfakes, identity politics, and influence campaigns were pervasive. Blackbird.AI’s Constellation Narrative Intelligence Platform synthesized data from across the internet to uncover narrative attacks aiming to shape public perception through narrative attack. Constellation aided in analysis, displaying online trends and discovering narrative attacks—based on an election campaign peppered with deepfakes and identity politics—that emerged from English- and Spanish-language conversations leading up to the June 2 election.

LEARN MORE: What Is A Narrative Attack?

Narrative Attack 1 – Distrust in the Public Agency Responsible for Organizing Federal Elections

Mexico’s electoral watchdog, the National Electoral Institute (INE), anticipated that 390,000 of the 1.4 million Mexicans living abroad would register to vote in the election by mail, online, or in person at a consular site. That’s more than double the number of registered expat voters in the 2018 election.

Limited to Mexico’s single six-year presidential term known as a sexenio, outgoing President Andrés Manuel López Obrador—commonly referred to as AMLO—proposed electoral reforms in 2022 and 2023 aiming to cut the INE’s budget, which would lead to a significant staff reduction and the closures of several local offices. AMLO had previously accused the INE of partisanship and argued that the budget cuts would make voting more efficient. The proposed plan to limit the INE sparked protests over concerns about Mexico holding a free and fair election in 2024.

After the INE rejected around 40,000 applications from expats attempting to register to vote, largely due to alleged inconsistencies with signatures, a dominant narrative emerged online. Mexicans criticized AMLO’s attempts to undermine the country’s democratic institutions. Conversations involved angry allegations that Mexico’s governing party, Morena, is corrupt, circulating the name of AMLO’s protégé, Sheinbaum.

Narrative Attack 2 – Leading Presidential Candidate Plagued by Deepfakes

In January, a video emerged online depicting Sheinbaum inviting Mexicans to make a small investment that would lead to substantial financial growth. She denounced the video on her social media platforms as a “complete fraud.” The goal of this pyramid-scheme deepfake was likely to undermine leading candidate Sheinbaum’s chances of becoming president. 

Additionally, the week before the election, an account impersonating a news outlet shared AI-generated audio of Sheinbaum allegedly admitting her campaign was failing in a key Mexican state.

Narrative Attack 3 – Identity Politics and Misinformation Campaigns

Following a 2019 constitutional reform called “gender parity in everything,” political parties must propose equal numbers of male and female candidates. The leading two of the three presidential candidates were women; Sheinbaum’s opponents included opposition candidate Xóchitl Gálvez and Citizen Movement Party candidate Jorge Álvarez Máynez. In a country that granted women the right to vote in 1953, Sheinbaum and Gálvez faced misogynistic narrative attacks leading up to the election. 

Appearance-based narrative attacks emerged alongside misinformation campaigns questioning Sheinbaum’s legitimacy as a presidential candidate. Last year, Sheinbaum, born in Mexico City and is Mexican by birth and nationality, posted her birth certificates to social media to quell claims that she was born in Hungary or Bulgaria, likely stemming from her grandparents having emigrated from Eastern Europe. In a predominantly Catholic country, these assertions also likely intended to stir up other narratives related to Sheinbaum’s Jewish background.

Narrative Attack 4 – From Virtual Violence to Real Bloodshed

Virtual violence in the form of deepfakes and misinformation campaigns was not the only form of brutality in the Mexican general election. Criminal groups seeking to influence the outcome of local elections had been threatening, kidnapping, and killing candidates in the lead-up to June 2, prompting other candidates to withdraw out of fear.

Around 750 victims connected with campaigns for nearly 20,000 local candidates have been murdered or threatened, painting a picture of the most violent election in Mexican history – especially in the state of Guerrero. Across Mexico, at least one candidate was attacked every four days, and local Morena Party candidates were statistically the most affected. 

Organized crime rates increased in Mexico across AMLO’s six-year presidency, despite his efforts to lower a continuously high murder rate through various initiatives, making crime a top concern for those voting in the June 2 election. 

The Way Forward

The prevailing dominance of social media and online campaigning, especially with the accessibility of generative AI tools, has paved the way to increasingly more misinformation-riddled elections. With these roadblocks, deciphering the truth and finding accurate context for online claims is more challenging than ever. 

Criminal groups and AI-generated audio and visual deepfakes both worked to influence the outcome of the Mexican general election, which has been yet another example of how narrative attacks threaten democratic processes. Even days after Sheinbaum won around 60% of the vote, narratives attacking her appearance and alleging election fraud continue to take shape and circulate.  

Blackbird.AI’s tools are here to protect you or your organization against these narrative attacks. Blackbird.AI’s Constellation Narrative Intelligence Platform identifies and analyzes harmful narratives as they emerge. Compass by Blackbird.AI provides crucial context as these narratives threaten the pursuit of factual information and the distortion of reality.

To learn more about how Blackbird.AI can help you in these situations, book a demo.