Photo: SERGIO LIMA/AFP via Getty Images
In what is perhaps the most influential philosophical treatise on the subject of free speech, On Liberty, John Stuart Mill famously penned, “If all mankind minus one were of one opinion, and only one person were of the contrary opinion, mankind would be no more justified in silencing that one person, than he, if he had the power, would be justified in silencing mankind.” The advent of social media and the ensuing arrival of viral disinformation campaigns, Russian bots, and deepfakes have called into question the logic and limits of Mill’s unqualified defense of free speech in democracies both here in the Western Hemisphere and in other parts of the globe. Brazil, for example, is in the midst of a countrywide experiment in which the federal government looks to become a key protagonist in the effort to combat the spread of disinformation and antidemocratic expression. This is happening against the backdrop of a particularly bitter presidential election that was followed almost immediately by a violent attack on its executive, judicial, and legislative seats of power.
For some, the events that played out on January 8 in Brasília demonstrated the need for social media companies and possibly even government authorities to monitor and ultimately censor erroneous claims of election fraud. Drawing a direct line from fallacious news stories shared via social media to political violence committed in the heart of the nation’s capital, social media companies and government authorities alike in Brazil are feeling the pressure to intervene as online content moderators. But as the Brazilian government contemplates the creation of additional mechanisms to crack down on disinformation and antidemocratic expression, the question to consider is whether such designs are likely to solve the problem they are meant to address.
Before outlining the specific steps that Brazil is taking to address the very real problem of disinformation, it is worth acknowledging that virtually every social media company has some type of content moderation policy. To be sure, many of these platforms quietly go about the business of eliminating certain forms of content—from spam and online nudity to direct threats of violence—without much ado. Having the necessary safeguards to first pinpoint and later expurgate such content makes sense, especially considering what these platforms would look like without content moderation policies. Accordingly, the issue at hand is not whether private social media companies have the right to create and enforce content moderation policies, nor is it whether governments have any legitimate right to regulate these platforms. The question is, rather, whether governments should be in the business of enacting speech restrictions that are content- or viewpoint-based.
In a meeting with foreign diplomats that took place last July in the Palácio da Alvorada, then-president Jair Bolsonaro voiced his concerns that the country’s electoral system was somehow vulnerable to manipulation and outside interference. As documented by a New York Timesreport, this was hardly the first time the former army paratrooper had trafficked in baseless claims of electoral fraud. His invocation of a possible intervention by the Brazilian armed forces, however, was particularly troubling given the country’s relatively recent experience with military rule.
During the run up to last year’s presidential election, social media users in Brazil were inundated with a barrage of vitriolic content in which candidates were accused of everything from cannibalism and pedophilia to paganism and devil worship. With an estimated 75 percent of the Brazilian population on Facebook, Twitter, Telegram, or TikTok, if not all four, these platforms have become a common way for Brazilians to share and receive news. In fact, more people in Brazil turn to WhatsApp for their daily dose of news than those that subscribe to legacy media outlets. While Brazil is hardly the only country to witness such dramatic changes in the surrounding media landscape, its acute reliance on a single application makes it particularly vulnerable to the deleterious effects of misinformation and disinformation.
Four months before the events of January 8 played out, Brazilian Supreme Court justice Alexandre de Moraes ordered the police to raid the houses of several influential businessmen who, in a private WhatsApp chat group, mused over the comparative advantages of a possible coup in the event that the country’s then-president Jair Bolsonaro lost the upcoming election. Having already directed social media companies to remove thousands of posts that he personally deemed to be pernicious, Moraes was later granted near unilateral authority to effectively determine which content was permissible to share online during the final few days leading up to the October vote. Such active involvement on the part of Moraes is illustrative of the Brazilian legal framework, which employs a civil rather than a common law system wherein judges are bestowed with increased investigatory powers. Members of the Brazilian judiciary have even gone so far as to issue nationwide moratoriums on certain social media applications. Telegram, for example, was suspended in March 2022 by Alexander de Moraes for ignoring orders to remove posts containing disinformation. Telegram was suspended again in April 2023 by a different judge for failing to provide law enforcement agencies with the user data of individuals belonging to neo-Nazi groups, including the alleged perpetrator of a pair of school shootings from last November.
Upon taking office on January 1, President Luiz Inácio Lula da Silva (Lula) went on to create two new departments, a working group, and a designated website to address the problem of disinformation. The newly minted Secretariat for Digital Policies has been tasked with combating online hate speech. The Office of the National Prosecutor for the Defense of Democracy, meanwhile, is responsible for combating disinformation with respect to public policy. The Brazilian judiciary is also considering ways to hold social media companies responsible for their perceived failure to prevent the circulation of conspiracy theories on their respective platforms. As it stands, Article 19 of Brazil’s 2014 Marco Civil law effectively shields these companies from liability "for damages resulting from content generated by third parties." Therefore, in the absence of a court order, any attempt to compel these companies to remove monetized content deemed to be outside the bounds of acceptable speech would require a change to the country’s regulatory framework.
The debate over disinformation in Brazil reached a fever pitch in recent weeks. At issue is the resurrection of the so-called Fake News bill (PL 2630) which, after having been approved by the Brazilian Senate back in 2020, lingered for years in the Chamber of Deputies as it went through a series of substantive changes. At one point, the bill proposed the creation of an autonomous government agency that would dictate which content social media companies should remove from their respective platforms. At present, besides holding social media companies responsible for monetized content published by third party users, the bill would also compel media platforms to inform law enforcement agencies of content that could possibly result in criminal activity. Full of vague terminology, the bill would also require social media companies to remunerate media outlets for content that is uploaded to their respective platforms. The imminent passage of PL 2630 occasioned a blitz of opposition campaign messaging from companies like Google and Telegram, who were determined to push back against some of the bill’s more controversial provisions.
The need to avoid a reprise of the January 8 riots in Brasília has understandably pushed decision makers in Brazil toward more speech-restrictive policies. The worry, however, is that in their haste to take action, what they view as being immediate solutions turn out to be little more than conductors of further distrust and division. More concerning still is the prospect of government authorities assuming the job responsibilities of independent fact-checkers, such as that which effectively occurred with creation of the disinformation awareness website.
Although freedom of expression is enshrined in Article 5 of the Brazilian Constitution, it is important to keep in mind that, as a general rule, Brazilians do not enjoy the same free speech rights as U.S. citizens. Lacking the explicit language included in the First Amendment to the U.S. Constitution, Brazilian law, much like that of Germany and other European countries, affords the federal government greater leeway to proscribe speech considered to be harmful. A principled opposition to content-based or conceptually vague speech restrictions within the Brazilian context, therefore, must find much of its basis, not in constitutional dictates, but rather in the wisdom gleaned from both liberal philosophy and the historical record.
There are other reasons to steer clear of this type of censorship besides those arguments detailing the provisional nature of truth. First, by narrowing what is considered permissible speech on a given social media platform, users obligated to take their online conversations elsewhere are liable to become even more extreme in their views. Drawing a clear distinction between ignorance and safety, other civil liberties experts have articulated the benefits of having an accurate understanding of what other people think, especially when it may be worrisome. Censorship is also a notoriously unwieldy instrument. In the past few years alone, accounts of anti-racist activists have been suspended for speaking out against racism, a Pulitzer Prize-winning photograph demonstrating the harsh realities of the Vietnam War has been removed, and content moderation policies have tended to disproportionately target members of the LGBTQ+ community. As one final word of caution, it is also worth considering the distinct possibility that those advocating for more constraints on speech today could be on the receiving end of those same restrictions tomorrow.
Brazil’s efforts to thwart the proliferation of disinformation and its various analogues will inevitably be understood within the broader panorama of free speech issues in Latin America and the Caribbean. Unfortunately for individuals like Alexandre de Moraes, a cursory glance at the state of free expression in the region will likely compel observers to be skeptical of Brazil’s attempts to stamp out disinformation. The Western Hemisphere is home to several governments whose leaders have responded to their own internal crises by cracking down on expression. Unsurprisingly, many of these stratagems were marshaled for nakedly partisan political ends.
In October 2020, the Sandinista-controlled Nicaraguan congress passed the so-called cybercrimes law that threatens those accused of promulgating fake news with serious jail time. Similar legislation was passed in Cuba following the unprecedented wave of protests that took place in July of 2021. Later that same summer, the Cuban government issued Decree-Law 35, which levies hefty fines on telecommunications providers if they fail to prevent the circulation of fake news or reports. In Venezuela, the illegitimate National Constituent Assembly approved the Law Against Hate, which carries penalties of between 10 and 20 years in prison to anyone who runs afoul of the government’s views on speech. The law was approved at the request of President Maduro in response to calls on social media platforms to organize the massive anti-government protests that took place between April and July 2017. Not to be outdone, the semi-authoritarian government of Luis Arce is currently pursuing a similar bill (draft law 305) that, according to the president of the National Press Association of Bolivia, threatensjournalists with prison time in order to “silence any opinion that may make the government uncomfortable” (author’s translation). That Lula—in the name of defending democracy—would seek to restrict a technology that has been so instrumental in attempts to rescue democracy elsewhere, is curious to say the least.
Government attempts to arbitrate truth are not completely foreign to the United States. It was in April of last year when the Department of Homeland Security announced the creation of a new governance board whose professed raison d’être was to help combat the spread of disinformation in the country. Slammed by critics as a real-life imitation of George Orwell’s Ministry of Truth, the board was ultimately mothballed. The United States is also dealing with its own internal struggles over the prospect of increased social media regulation. It was just last week when the U.S. Supreme Court, in two separate cases, unanimously found that neither Google nor Twitter had been complicit in perpetuating acts of terror by hosting ISIS-related content on their respective platforms. In its two narrowly tailored rulings, the court managed to sidestep the floated issue of eliminating a key provision of the 1996 Communications Decency Act (CDA). At present, Section 230 of the CDA shields web platforms from potential liability over third-party content. Recognizing that such protections are a boon to online expression, civil liberties watchdog organizations such as the Electronic Frontier Foundation have detailed the negative consequences that are likely to materialize should the court ultimately decide to completely do away with Section 230. Others, like former Facebook employee turned whistleblower Frances Haugen, have argued that amending Section 230 is the only way to rein in profit-hungry social media companies that allegedly elevate divisive content to drive further user engagement even at the expense of exacerbating polarization. As is so often the case, there are principled arguments to be made on both sides of the debate. The implications for free expression that are inextricably tied to this issue, however, indicate that it is best to err on the side of more speech, rather than less. Although last week’s rulings were undoubtedly a huge win for the tech industry, the heated debate surrounding Section 230 can be expected to continue well into the future.
Compared to other attempts to take on the extant challenges of the twenty-first century, Brazil’s principal response to the problem of disinformation, namely compelling social media companies to remove unsavory posts as they appear, is haphazard. Much like climate change, mass migration, and the drug war, the Gordian Knot of disinformation will invariably require solutions that go beyond the level of superficial stopgaps. This means taking steps towards addressing deficiencies in areas like civic education and media literacy, not to mention doing the heavy lifting of restoring trust in public institutions.
More than anything else, the arguments presented here are intended to give both policymakers and voters in Brazil something to reflect on as they consider their options. It is true that uninhibited speech, much like censorship, presents its own set of trade-offs and unique challenges. If the past is any indication of what to expect moving forward, however, the costs associated with the latter vastly outweigh those associated with the former. In the face of creeping illiberalism and authoritarian power grabs, the citizens of democratic countries simply cannot afford such shortsighted thinking. Instead, the path to addressing these challenges should arc toward inclusive, open, and robust debate.
Christopher Hernandez-Roy is the deputy director and senior fellow at the Center for Strategic and International Studies (CSIS) in Washington, D.C. Michael McKenna is an intern with the Americas Program at CSIS.
The authors are grateful to Carlos Baena, an intern with the CSIS Americas Program, for his contributions to this commentary.
