The stakes may not be much at stake in Brazil’s elections. However, generation platforms are not, as expected, to cope with the moment.
More than 120 million Brazilians voted in the first round of elections on Oct. 2. A runoff on Oct. 30 will elect Brazil’s next president.
This election is likely to be a critical check on democracy and the rule of law in the country, with consequences that go beyond its borders, given Brazil’s extent and influence. In recent years, President Jair Bolsonaro, who is running for re-election, has sought to undermine accepting as true in the electoral system, claiming, without offering any proof, that it is unreliable. On October 17, just two weeks before the momentary round, Bolsonaro again questioned the reliability of the electoral system.
As we have noticed in elections around the world, social media platforms and personal messaging apps have the de facto public space of campaigns and public debates in Brazil. Technology platforms have a duty to respect human rights. This includes the right to participate in democratic debates. elections.
Brazilian civil society has warned that it opposes the dissemination of incorrect information related to the elections. And in February, the Superior Electoral Tribunal signed memoranda of understanding with Twitter, TikTok, Facebook, WhatsApp, Google, Instagram, YouTube and Kwai in a bid to stop incorrect information. Information on the electoral process. But the platforms have largely failed in their responsibilities. And unsurprisingly, the spread of electoral disinformation threatens to undermine Brazil’s democratic process.
On election day, and while the counting of votes was taking place in the following days, messages and videos containing incorrect information and allegations of electoral fraud began to circulate on social networks. WhatsApp and Telegram teams with supporters of President Bolsonaro circulated messages saying that “if Bolsonaro did If we did not win in the first round, it would be because the elections were fraudulent. “According to the Superior Electoral Tribunal, reports of disinformation are particularly higher compared to the first round. The TSE reported receiving 5,869 court cases in the first 11 days of the resumption of the crusade, which represented almost a portion of the court cases won by the crusade at that time. Complaints increased by 1,671% compared to the 2020 municipal elections.
People with thousands of fans claimed the count was fraudulent. That tweet, for example, reported that the “trap” initiated a 15-minute break from the electoral tribunal, hinting that it would favor Bolsonaro’s opponent. The tweet is still online, has more than 30,000 likes and contains no links to election management data, official election effects or other accurate data. Other tweets remain online that explicitly claim election “fraud” without any credible data, and do not include labels or links to authoritative data.
There are 3 key dimensions to this problem:
First, Brazilian civil society teams noted that no other platform, apart from Twitter, has a policy aimed at provoking calls contrary to the democratic order or interference in the nonviolent transmission of force that do not explicitly call for violence. This means that platforms can only be used to organize and publicize anti-democratic movements on the occasion of a post-election institutional crisis.
Second, when it comes to voter fraud, platforms have very different policies, and even the most powerful ones don’t apply. Meta’s publicly available policies, for example, prohibit the promotion of paid content that alleges voter fraud, but do not address similar unpaid content.
An investigation by the Netlab of the Federal University of Rio de Janeiro found that Meta allows certain content to circulate on Facebook and Instagram, even after it has been classified as electoral disinformation by independent fact-checkers, in some cases without labels. Most of the study’s messages argued that electronic voting machines are unreliable and that it would be easy to manipulate elections. The study also found that the profiles of those who posted the most incorrect information about elections were political.
Another Netlab study documented at least 4 instances of paid classified ads on Meta that contained unsubstantiated claims questioning the reliability of the vote count.
Twitter has a policy on civic and election integrity. But as the tweet repeating false allegations of voter fraud shows, the company doesn’t enforce the policy correctly. YouTube’s election integrity policy is limited to false allegations of voter fraud in past elections.
An investigation through the knowledge analysis consultancy Novelo Data published through Folha de São Paulo found that from October 2 to 15, 16 live broadcasts and 137 videos were published on YouTube with allegations of fraud without evidence. In total, the content had at least 3. 3 million views. Much of this content is still available. Telegram is an outlier. It has published a policy to combat disinformation and attacks on democracy, and has fulfilled its March commitments. Telegram and WhatsApp are said to be the main platforms used to spread incorrect information about the electoral process, with teams and channels with tens and thousands of subscribers committed to spreading this narrative. Closed messaging platforms pose different demanding situations to detect and eliminate incorrect information.
Third, tech platforms have a history of twisting their regulations for tough actors and allowing politicians to get away with speech that violates their policies. Access to what politicians say is, in fact, holding leaders accountable. But being more permissive of hardliners can allow them to incite violence on those platforms, or cause other harm, with little consequence.
Since politicians have more influence on society than people, their statements are, in fact, much more likely to cause harm. Many of the biggest spreaders of fake news in Brazil are at the highest levels of national politics. The formula has been propagated through President Jair Bolsonaro and his sons, who are also politicians.
Human Rights Watch wrote to Google/YouTube, Meta, Telegram, and Twitter to inquire about their efforts to mitigate the lack of electoral data in Brazil’s elections. No company responded directly to our express questions, but Meta said it has created teams that advertise reliable data and label elections. -related messages, established a direct channel for the Superior Electoral Tribunal to send potentially harmful content to it for review, provides access to information about content removed from Facebook and Instagram during the campaign, and continues to negotiate heavily with the Brazilian government and investigators.
Twitter’s written reaction highlighted its civic integrity policy, which uploads misleading data that would possibly have an effect on the integrity of civic processes, states that it will remove and/or load labels and context to misleading and destructive tweets and decrease their visibility. Twitter notes that opinions, other perspectives and interpretations are not necessarily subject to prosecution under its policy.
To fulfill their responsibilities, platforms must promptly comply with the orders of the Supreme Electoral Tribunal to eliminate electoral disinformation. They also fill gaps in their policies and implement them to ensure they respect Brazilians’ right to participate in democratic elections. This requires resources good enough to protect electoral integrity and civic discourse and to be transparent and accountable in their actions. The platforms also keep politicians in a top popular position for discourse that can incite violence or spread destructive disinformation that can undermine the democratic process.
Specific efforts to counter electoral disinformation are mandatory but insufficient. More broadly, platforms want to address their chronic lack of investment in the security of users around the world. of non-public knowledge to sell access to people’s care and are largely designed to prioritize their “commitment” to human rights.
* An edition of this article published in the LOU.