On October 2, Brazilians will head to the polls in a presidential election that will be among the most decisive in recent memory – not only for Brazil, but for Latin America. As the race between populist President Jair Bolsonaro and socialist former president Luiz Inácio Lula da Silva enters its final weeks, Lula, as he is known, continues to lead in the polls, but the contest is tight. The campaign has been highly polarized and divisive. As Global Fellow at the Brazil Institute Nick Zimmerman said in a recent podcast discussion, the elections will be an expression of “the state of being of global democracy, where democracy stands in the Americas and in the Global South.”
For months, President Bolsonaro has questioned Brazil’s voting systems, casting doubt on their security and claiming they are vulnerable to vote rigging, although no evidence of this exists. The Brazilian military has reinforced the president’s arguments, claiming that hacking software could be installed on voting machines. Bolsonaro has also suggested the need for a parallel vote-counting system that would involve the armed forces, an idea supported by a naval commander.
Since the start of Bolsonaro’s presidency in 2019, Brazil’s democratic institutions have been under stress. And it’s not overstating it to say that new technologies have played a central role in the country’s democratic backsliding. Brazil was once considered a pioneer of digital rights and internet freedom, thanks to its 2014 Marco Civil Law of the Internet. Providing guidelines for all levels of government, the law states that the use of the internet in Brazil must be based on principles of freedom of speech and communication, protection of privacy, protection of personal data, net neutrality, and preservation of stability and security.
In 2022, however, the law’s ambitious principles are under siege. Indeed, President Bolsonaro’s rapid rise to power can be seen as an example of some of the negative impacts of new technologies on democratic processes and institutions.
When he ran in the 2018 presidential election, Bolsonaro was a retired military officer and a little-known political figure who had been a member of the lower chamber of Brazil’s Congress since 1990. Few at that time could have predicted the dazzling rise of this right-wing populist.
Much like his North American counterpart President Donald Trump, Bolsonaro took advantage of the messaging power of social media platforms, whose business models amplify disinformation and inflammatory speech. Since the beginning, Bolsonaro’s campaigns have used social media to spread false information about the elections and to attack journalists such as Patrícia Campos Mello, representatives of Congress, and non-governmental organizations that dared to criticize him.
Once elected, Bolsonaro followed the same path, this time with a state apparatus behind him. Freedom House’s Freedom on the Net report now ranks Brazil as “partly free,” largely due to the president’s and his allies’ online harassment and monitoring of journalists, civil society activists and political opponents. Online threats have been accompanied by an increasing number of criminal investigations against critics, according to the report. Carlos Bolsonaro, one of the president’s sons, along with advisers of the president, established a secret “office of hate.” This online operation amplifies pro-Bolsonaro news websites, YouTube vloggers and social media accounts that spread propaganda, disinformation and hate speech.
At this writing, the president himself has 8.9 million followers on Twitter and 14 million on Facebook, making him the third-most-followed national leader on the latter platform. At the height of the COVID-19 pandemic, which killed more than 680,000 Brazilians, disinformation about the virus and disease flooded Brazil’s information ecosystem, circulating more within political groups aligned with the far right and Bolsonaro.
Bolsonaro’s own role in spreading disinformation led a 2021 Brazilian congressional panel to recommend that the president be charged with crimes against humanity for his response to the pandemic. In 2022 the country’s federal police followed suit, calling for the president to be charged with spreading falsehoods.
Since 2018, Bolsonaro has also used his presidential powers to pass several decrees and laws that have steadily eroded data privacy and freedom of expression online. The Brazilian Law of Freedom, Responsibility and Transparency on the Internet introduced in the Federal Senate of Brazil on May 13, 2020, sought to restrict and penalize legally protected speech, making it a crime to create and share content that posed a serious risk to “social peace or to the economic order” — a vague definition. The Global Network Initiative writes that the bill’s provisions also included “requiring users to register for social media and private messaging services using government identification documents, together with … data retention requirements.” In June 2022, the Brazilian Congress rejected the fast-tracking of the bill and its future is now uncertain.
And this has not been the government’s only attempt to regulate freedom of expression online. In September 2021, irked by YouTube’s decision to remove videos posted by his office that spread disinformation about the coronavirus, Bolsonaro signed a provisional executive measure, Measure 1068/2021, on content moderation. The measure temporarily banned social media platforms from “adopting criteria of moderation or limiting of scope of dissemination of content” that might imply censorship of political, ideological, scientific, artistic or religious content — including medical and political misinformation. The government claimed on Twitter that it was “taking the global lead in defending free speech on social networks and protecting the right of citizens to freedom of thought and expression.” The provisional measure was rejected by Brazil’s Supreme Court and Congress one week later.
During a now infamous cabinet meeting on April 22, 2020, Bolsonaro expressed his disdain for the work of existing intelligence agencies and admitted that he maintained his own information service.
Thankfully, Brazilian civil society organizations are keeping track of the executive branch’s attacks on digital rights. The Brazilian non-profit civil society organization Data Privacy Brasil (DPBR) has coined a term to describe the phenomenon: “techno-authoritarianism.” Techno-authoritarianism, the organization states, “can be used to explain the processes of expansion of state power whose objective is to increase surveillance and control capacities over the population, by violating individual rights or significantly increasing the risks of violation of fundamental rights.”
Rafael Zanatta, the director of the DPBR, explains that Brazil’s techno-authoritarianism includes not only attacking critics on social media, but also centralizing personal databases; abusive sharing of data with intelligence and public security agencies; expansion of surveillance, including through facial recognition; and the development of new information and intelligence systems, sometimes with the involvement of the armed forces.
Indeed, regulated by the Brazilian National Civil Identification Law, Brazilian citizens’ personal data, including sensitive biometric data, has been increasingly centralized and made accessible to executive and legislative powers. On March 18, 2020, Bolsonaro issued a decree “allowing the government to share classified personal data with other public agencies without the need for express authorization from citizens.”
A few months later, the Brazilian Intelligence Agency (ABIN), which is linked to the federal government, and the Brazilian Federal Data Processing Service agreed to share with each other the personal data of all Brazilians with a National Driver’s License, giving access to a trove of personal information. The government was eventually pressured to revoke the agreement after the Brazilian Socialist Party filed a lawsuit with the Federal Supreme Court. But such centralization, sharing and free flow of data clearly violate data protection principles under Brazil’s 2020 General Data Protection Law.
Of particular worry is the establishment of parallel intelligence and information channels. To DPBR director Zanatta, it has always been clear that Bolsonaro would try to dismantle how public security and intelligence operations are conducted in Brazil. A former member of the military, Bolsonaro has seemed unwilling to trust existing institutional structures. With the help of his allies, he has tried to meddle in federal police business and attempted to gain access to secret intelligence reports.
During a now infamous cabinet meeting on April 22, 2020, Bolsonaro expressed his disdain for the work of existing intelligence agencies and admitted that he maintained his own information service: “Information systems: mine works. Mine, particularly, works,” he said. It is important to note that the president’s inner cabinet is highly militarized: in 2020, seven members of a 20-member cabinet were military men, and his vice-president, Hamilton Mourão, is a retired general. There is just one woman in his cabinet.
Another worrying development has been the creation of secret dossiers of public security officials and intellectuals. In July 2020, the Brazilian Ministry of Justice and Public Security, through its Secretary of Integrated Operations (SEOPI), whose actions are not monitored by Brazilian courts because it is considered an intelligence service, produced dossiers on 579 people it regarded as “threats of institutional democratic destabilization in our country.” These dossiers contained photographs, addresses and social media accounts, and were shared with the president’s cabinet, ABIN, the federal police and the army’s information centre. While gathering of such dossiers was eventually suspended and deemed unconstitutional, the Inter-American Commission on Human Rights said these steps were reminiscent of the days of Brazil’s military dictatorship,which lasted 21 years, from 1964 to 1985.
Today, some Brazilians are taking up the fight and going to court to confront techno-authoritarian tactics. In conversations, Rafael Zanatta and Artur Pericles (a Wikimedia Fellow at Yales Information Society Project, who has been researching content moderation) have told me the country’s courts have become the battlefields of resistance. Pericles says that surprisingly, even Brazil’s left-leaning parties, including Lula’s Workers’ Party, rely on courts to side with tech platforms, in an attempt to fight back some of Bolsonaro’s measures. It is quite rare to see the left side with big business.
Organizations such as the DPBR have formed a consortium focused on data protection and fundamental digital rights. In 2019, when the Brazilian Superior Electoral Court sought to draft new regulations, including one dedicated to regulating electoral propaganda on the internet, the Internet Lab, an independent research centre based in Brazil that aims to foster debate on law, technology and internet policy, proposed several resolutions related to data privacy rights and freedom of expression, some of which were accepted by the Superior Electoral Court. Zanatta also told me that a team at the consortium is currently monitoring any kind of automated mass messaging during campaigning, which is illegal according to the Brazil Supreme Federal Court but was nonetheless used by Bolsonaro’s campaign in the 2018 presidential race.
In Brazil, the lack of transparency regarding content moderation, including during elections, continues to be big tech’s major flaw.
The rise of techno-authoritarianism does not bode well for Brazil’s October elections.
To spread their message, the president and his supporters are replicating digital tactics used in the 2018 presidential elections. As part of a study at Columbia University, Brazilian journalist Patrícia Campos Mello is currently monitoring Telegram and WhatsApp groups as well as right-wing propaganda websites that present themselves as impartial and credible news outlets. Campos Mello says Bolsonaro’s campaign, his political allies, extreme bloggers and hyper-partisan websites feed off one another and amplify the incumbent’s message. Combined with the claim of vote rigging and an untrustworthy electoral system, three main messages emerge: first, mainstream media is not to be trusted; second, polling is not to be trusted and Lula is not leading in the polls; and third, Brazil’s current economic woes are not President Bolsonaro’s fault. According to Campos Mello, a separate news ecosystem has been created that energizes and mobilizes the core base of Bolsonaro supporters.
Social media platforms are not taking the lead to prevent election disinformation from spreading on their platforms. A study published by Global Witness in August 2022 shows that Facebook has failed to tackle election disinformation about the upcoming elections in Brazil, despite Meta’s policies on safeguarding election integrity. Global Witness submitted 10 Brazilian Portuguese-language ads to Facebook containing election-related disinformation: Facebook accepted all of them.
Rosie Sharpe, one of the authors of the Global Witness report, shared with me the ads submitted to Meta. They contained false slogans such as “Voting is voluntary now” (voting is mandatory), false information about the day of the election, false information about voting by mail (it does not exist in Brazil), information discrediting the electoral process and voting machines, and even slogans to the effect that “we must undertake a staunch fight in favour of the printed ballot, so that what happened to the United States does not happen to Brazil.”
In Brazil, the lack of transparency regarding content moderation, including during elections, continues to be big tech’s major flaw. In February 2022, Patrícia Campos Mello criticized agreements released by the Brazilian Superior Electoral Court and internet platforms, saying that they “fall far short of the electoral policies adopted by companies in the United States.” The agreements signed with eight platforms (Twitter, TikTok, Facebook, WhatsApp, Google, Instagram, YouTube and Kwai) sought to avert election mis- and disinformation, but Campos Mello noted that they have remained vague about how they will respond if election results are contested or if violence is incited on their platforms: “Except for Twitter” — which banned political ads globally in 2019, and has either made fake election content subject to removal or labelling or reduced its reach — “none of the companies — Google/YouTube, Facebook, TikTok, and Kwai — currently specify how they will react in the event of a massive campaign of electoral disinformation,” she writes. Telegram, which is used by 50 million Brazilians, refused to cooperate with the Superior Electoral Court until the Federal Supreme Court (the top electoral authority) considered banning the messaging app altogether. In May, the Superior Electoral Court and Telegram finally reached an agreement to combat disinformation. Many loopholes remain. In early August, Campos Mello wrote that big tech companies still had not shared concrete information about the moderation teams responsible for monitoring content during the election period.
Heading into the pre-election period, the situation is deeply worrying and the outcome very uncertain. Should Lula win, Jair Bolsonaro is unlikely to accept the election results, as he has already suggested. Campos Mello fears such a reaction could lead to a January 6 type of insurrection, and this is not a wild suggestion. In August, the Supreme Federal Court ordered search warrants against eight business leaders who allegedly participated in talks on WhatsApp about a possible coup should Lula and his party be elected. If Bolsonaro wins, there’s a real prospect of an even greater erosion of democracy, with checks and balances weakening further. The next president will be empowered to appoint at least two new justices to Brazil’s Supreme Court.
Whatever the outcome, techno-authoritarianism has clearly found its roots in Brazil, eroding this once-burgeoning democracy. The outcome of the election will tell us which trajectory Brazil will take.
