Days after the first round of voting in the Brazilian election, the Atlantic Council’s Adrienne Arsht Latin America Center and Digital Forensic Research Lab (DFRLab) hosted the conference call “Brazil’s Vote: The Role of Disinformation in the 2018 Elections” to discuss the impact of disinformation and misinformation on Sunday’s results.
The Atlantic Council’s #ElectionWatch Latin America initiative has identified, exposed, and explained disinformation and the spread of misinformation in this year’s elections in Colombia, Mexico, and Brazil. In Brazil, findings reveal that disinformation and misinformation circulated across Latin America’s biggest democracy as voters headed to the polls in an extremely polarized environment.
Graham Brookie, director of the Atlantic Council’s DFRLab kicked off the call by stressing the mission of the Adrienne Arsht Latin America Center and DFRLab to create digital resilience. He noted the importance of understanding key differences between disinformation and misinformation, the first being the spread of untruthful information with intent, and the latter being the spread of misleading information.
Roberta Braga, associate director and lead on Brazil at the Atlantic Council’s Adrienne Arsht Latin America Center, moderated the discussion between Luiza Bandeira, digital forensic researcher at the DFRLab and Adrienne Arsht Latin America Center; Daniel Bramatti, president of the Brazilian Association of Investigative Journalism (ABRAJI) and coordinator of the Comprova Project (a network of journalists in Brazil); Danilo Carvalho da Silva, researcher and analyst at the Department of Public Policy Analyses at Fundação Getúlio Vargas (FGV DAPP); and Leonardo Cazes, editor of Fato ou Fake, an initiative under the umbrella of Brazil’s Jornal O Globo. “From electoral fraud allegations, to the buying and selling of likes and shares, to distorted photos and videos, and a migration of Brazilian users to GAB, we’ve seen examples of disinformation spreading like wildfire online for months,” Braga commented. The issue, she noted, will continue to impact democratic discourse ahead of the second round of elections on October 28.
In highlighting key findings of her Atlantic Council research, Luiza Bandeira detailed that websites spreading distorted news often outperform traditional news sites. Usually one-sided and exaggerated, the goal of these websites is to influence voters. The spread of misleading news usually happens through Facebook and Twitter, with WhatsApp having a very large reach in Brazil. For example, after Jair Bolsonaro was stabbed during a campaign event earlier this year, Bandeira noted that although fact-checking organizations were highly productive in debunking false claims, their articles “were not reaching the [audiences spreading] them.”
Contributing to the susceptibility of disinformation in the election is the decreased trust in traditional media in Brazil. Daniel Bramatti remarked that the intentional targeting of news outlets during the elections foments polarization and creates an environment of mistrust. He sees this mistrust as extremely dangerous, saying that “the problem is not only that we are not reaching some part of the Brazilian audience. When we do, they react as if we are lying.” Leonardo Cazes, editor at Fato ou Fake, complemented Bramatti’s argument by stating that “when [people] don’t trust any [kind of institution], […] they can trust anyone or any kind of information that they have in their hand, and in their cell phones.” This makes the spread of disinformation more predominant.
With a large portion of society receiving most of their news from social networks, Cazes noted that tracking disinformation, particularly within WhatsApp, can be extremely difficult. The small group structure of the app can be a hub for the exchange of false news. Some organizations, like Brasil Paralelo (Parallel Brazil), take advantage of the platform’s shareability to advance their arguments, making it even harder for fact-checking organizations to verify and debunk distorted information within the platform.
Danilo Carvalho da Silva also highlighted the problems that come with the use of bots during the presidential campaign. The first week of campaigning saw 10 percent of all Twitter interactions coming from bots, FGV DAPP found. Their interaction on the same platform doubled the week before the first round of elections and on the day of the vote. Carvalho explained that “while the bots in pro-Haddad groups were acting mainly as amplifiers of pro-Haddad content… bots in pro-Bolsonaro groups spread a lot of disinformation.” While this has been noted prior to the election, Carvalho affirmed it was done more effectively and proactively during the cycle.
The panelists also expressed concern over false claims of electoral fraud, which circulated widely during and immediately after the first round of voting. In Brazil, where distrust in institutions in general is at an all-time high, false claims of electoral fraud can delegitimize the democratic process, and throw into question the results, the panelists affirmed.
In looking to potential solutions to the spread of disinformation in Brazil, Roberta Braga argued that in addition to efforts made by tech companies, media, and the government, every individual also has a role to play in actively verifying information. Simple steps, like double checking information on Google before sharing, could reduce the spread, Bramatti suggested. Cazes added that individuals must think twice about whether the information is reasonable, particularly videos and photos, and should share the debunking articles more often. To conclude the discussion, Brookie affirmed that the “challenge of disinformation is truly collective, and any solution is going to require collective action that includes government and social media and media companies.”