Social Media Platforms Don’t Have Policies to Prevent Insurgency in Brazilian Elections, Says a Report
According to the newspaper Folha de São Paulo, a document signed by 116 entities, warns that of all the social media platforms present in Brazil, only Twitter has policies to prevent calls for uprising against the democratic order or interference in the transmission of power in Brazil that do not explicitly call for violence.
The document will be released this Friday (16) and sent to the companies.
According to the entities, in a scenario of institutional crisis during or right after the October elections, the platforms may become an environment of organization and promotion of antidemocratic actions, as it happened in the US with the invasion of the Capitol on January 6, 2021.
Most platforms prohibit content with explicit incitement to violence. But they do not have moderation rules for less obvious cases.
For example, if a politician makes a video raising doubts about the vote count or posts content encouraging his supporters to monitor the elections in a certain polling station, this is not an explicit encouragement to violence, but it can result in rioting; or if, right after the election, with contested results, officials or politicians use social media to say that the Constitution allows an intervention of the Armed Forces.
This type of posting is not contemplated in the rules of the platforms and can be a trigger for aggression and turmoil, the entities warn.
The only company whose rules address this type of situation is Twitter, which provides for removing or labeling misleading information about outcomes, such as “controversial claims that may call into question faith in the act itself, such as unverified information about voter fraud, vote tampering, vote counting, or certification of results.”
Twitter vetoes misleading claims such as celebrating victory before the election results have been certified, inciting illegal conduct to prevent the implementation of election results.
“The advancement of democratic commitment in the platforms’ policies is fundamental for them to assume their responsibilities to the rule of law,” says Francisco Brito Cruz, executive director of InternetLab and member of the group of organizations that made the letter.
“Almost 15 days before the first round and even with advances, several questions still remain unanswered, or with insufficient answers. It is unacceptable that Telegram, for example, does not have a public policy for moderating content that attacks the integrity of the elections.”
According to the document, Telegram still does not have a policy to combat disinformation and has not met the commitments made with the authorities.
Earlier this year, Minister Alexandre de Moraes blocked the app in Brazil and made several demands — among them, to inform him of the measures it was taking against the spread of fake news.
In July, the entities released a report with requests for more effective measures, but most of the suggestions were not implemented by the companies.
“In general, the platforms have policies to combat disinformation against the integrity of the electoral process, but disinformation against candidates continues with few restrictions,” says the document, signed by organizations such as SBPC – Brazilian Society for the Advancement of Science, Coalizão Direitos na Rede, Coalizão Negra por Direitos, Conectas Direitos Humanos, Abraji, Pacto pela Democracia, Observatório Político e Eleitoral, Intervozes, and Oxfam.
In February, Twitter, TikTok, Meta (owner of Facebook, WhatsApp, and Instagram), Google (owner of YouTube), Kwai, and Telegram signed agreements with the TSE (Superior Electoral Court) with commitments to combat electoral disinformation.
According to the document, the agreements are “insufficient to limit disinformation in the electoral process.” “Still, their implementation has advanced across the platforms, with the exception of Telegram.”
In the wake of the Capitol Hill invasion, Meta earlier this year adopted a rule to restrict accounts of public people (politicians, candidates, people with more than 1 million followers) during civil unrest. But accounts will only be suspended if they incite or exalt violence.
The document cites that Facebook/Instagram (Meta) banned, as of August 16, 2022, political ads that question the legitimacy of the Brazilian elections after an experiment by Global Witness detected a flaw in the filtering process of boosted advertising that violated the platform’s rules.
“Although the policy update is considered a relevant step, a monitoring conducted by Netlab/UFRJ indicates that new violating political ads continue to go undetected and barred by Facebook’s system,” it says.
The report praises the expansion of the scope of Facebook’s ad library to include the same themes considered sensitive in the United States. But it criticizes the election integrity policy, saying it is the least comprehensive of all platforms.
Facebook “does not classify as disinformation, for example, posts that contain unfounded allegations of voter fraud, which are not necessarily considered violative of the election integrity policy. Nor does it limit direct attacks on the integrity of the election process.”
Twitter, meanwhile, despite being the only platform with a clear policy to address misinformation about the electoral process, does not act on misleading information about candidates unless it violates other platform rules, according to the document.
YouTube this year began including in its election integrity policy content false allegations of vote fraud in the 2014 and 2018 Brazilian elections, as well as incorrect information about candidacy requirements and information stolen by hackers.
But, according to the document, it “does not cover cases in which it is alleged, in an unfounded manner, that the electronic voting system of the upcoming election process, such as that of 2022, or any election using electronic ballot boxes, will be or has been rigged.”
Nor is there a specific policy to prevent calls for uprisings against the democratic order or interference in the peaceful transmission of power that does not explicitly call for violence.
“Despite stating that it removes “harmful conspiratorial” and “incitement to interference in democratic processes” content, the platform does not interpret its rules for cases in the Brazilian context,” the study says.
RECOMMENDATIONS FOR THE PLATFORMS
Meta (Facebook and Instagram)
- Categorize as disinformation posts that contain unfounded allegations of voter fraud.
- Adopt specific policy to prevent calls for uprisings against the democratic order or interference with the peaceful transmission of power, even if there is no explicit call for violence.
- Adopt specific policy to prevent calls for uprisings against the democratic order, even if there is no explicit call for violence.
- Expand its policy to consider as disinformation subject to action false allegations that the electronic voting system of the 2022 election process is illegitimate and subject to fraud.
- Immediately reconfigure the recommendation system to ensure political neutrality in the content recommended to users.
- Keep banned ads in its report.
- Include not only links but also the content of the ads, for review by independent researchers.
- Adopt specific policy to prevent calls for uprisings against the democratic order.
- Establish a policy on the transparency of political advertisements.
- Establish and effectively enforce a policy to combat disinformation about elections on your public channels and groups.
- Enforce its policy of not allowing the use of the app for the circulation of misleading and false information, based on user complaints.