Lori Regattieri is a social-environmental-climate justice tech and movement builder, and is presently Senior Fellow in Trustworthy AI at the Mozilla Foundation.
After the 2022 Brazilian General Election and the violent riots led by supporters of far-right former President Jair Bolsonaro on January 8th, 2023, the acute nature of Brazil’s political crisis highlights how tech and media power asymmetries have real consequences for the civic information space, and thus on the most marginalized groups in Brazil. Amid global concern about the necessity of accessible digital public infrastructure to foster a healthier internet, demands for artificial intelligence worthy of trust, and the contemporary debate about reparations in tech and media, Brazil’s case exposes the necessity of policies grounded in evidence-based research from the Global South and informed by participatory mechanisms.
Bolsonaro’s campaign and presidency represented a cycle of propaganda that took advantage of a chaotic digital public sphere, fueling distrust in institutions and journalism. In this context, Bolsonaro and his supporters spread climate denialism while his public announcements against forest protection ultimately supported criminal activity in the Amazon. He and his government attacked human rights activists, environmental defenders, and indigenous leaders who attempted to stand in the way.
Since the January 8th storming of the country’s Congress, Supreme Court, and presidential palace in Brasilia, initial investigations have turned up evidence that the far-right propaganda machine found financial support in part from the agribusiness sector– the logging, illegal mining, and land grabbing groups that represent a significant part ofBolsonaro’s base. But theirs is not the only financial interest. For years prior to the 2022 Brazilian General Election, civil society and research labs repeatedly pressured Google and Meta to curb the use of their platforms and ad systems to nurture and sustain the far-right disinformation ecosystem.
To understand the information crisis in Brazil during the 2022 Elections and the subsequent violent attack against democracy on January 8th, it is necessary to unpack the participatory networked propaganda ecosystem where precarious work takes place, commercial media concentration, and how AI systems obscure platform decisions – harming marginalized communities and suppressing urgent matters from public debate, such as environmental racism, climate reparations, and territorial and indigenous rights. In order to address the information crisis in Brazil and produce a healthy public discourse that can help heal its democracy, it is necessary to take a critical approach to media and tech, backed by research and with all of this context in mind.
Brazil’s long road to January 8th connects cross-cutting issues in media and technology
The current media and information crisis has its roots in the Brazilian dictatorship era, from 1964 – 1985. After Brazil’s transition from military rule to a democracy, institutions such as the media, government, and the judiciary system that were complicit in the dictatorship acted to suppress meaningful access to the truth of what happened in those years. For example, it was only in 2011 that Brazil set up a truth commission to investigate human rights abuses committed during the period of military rule. In 2013, the major Brazilian corporate media outlet, Globo apologized for supporting the 1964 military coup. For context, it’s well known that powerful media conglomerates set the agenda for public conversations without offering alternatives that highlight historically marginalized voices. An important factor in Brazil’s information crisis is the lack of diverse perspectives reflected in Brazil’s current media landscape. Overall, since the end of military dictatorship, access to information and the appropriate treatment of the violent legacy of military rule in mainstream media has often been undermined due to the connections still in place in Brazilian civic and democratic institutions.
Today, Brazil has a classic hybrid media ecosystem in which a few oligarchic families and religious groups own the broadcast mainstream media. This creates asymmetries, such as narratives promoting neoliberal economic reforms broadcast without any potential for counternarratives to reach the public. For instance, very few spaces exist to voice the environmental concerns of indigenous peoples and local agricultural communities that oppose agribusiness propaganda. In this media environment, police brutality against Black communities is often depicted as ‘successful policy,’ further reinforcing structural racism. Meanwhile, social movements face staggering barriers due to false representations that muddy debates over justice and dignity. The struggle for land reform has long been covered in mainstream media under the premise of public policy misconceptions about territorial rights, as well as misleading content amplifying moral panic about Brazil’s Landless Workers Movement (MST).
As highlighted by University of Pennsylvania Annenberg School for Communication Professor Victor Pickard, systemic failures enabling the emergence of Bolsonaro and former U.S. President Donald Trump, as well as election, pandemic, vaccine, and climate related conspiracies, were “decades in the making”. In Brazil, Janaine Aires and Suzy Santos investigated the political economy of the Brazilian telecommunications system through the project “Electronic Coronelismo: asymmetric dynamics of power and negotiation”. The authors showed that the monopoly power in broadcast media and the distribution of local news affiliates historically affected the relationship between the audience and the perception of public policies in Brazil. The structure of the Brazilian media system kept the liberal aspect of distribution of public media concessions without considering the social and educational aspects of communication in the country’s 1988 Constitution. It’s important to highlight that access to information is a human right in Brazil. Unfortunately, every time serious discussions about structural reforms are proposed, corporate media reacts quickly to shut down conversations about the systemic roots of the information crisis.
The Intervozes report “Media Ownership Monitor Brazil” reveals the appalling media power concentration that still prevails in the country. According to the report, the four main Brazilian media groups concentrate more than 70% of the television audience. Considering key categories such as audience concentration, regulatory safeguards (ownership concentration and transparency), cross-media ownership, and political control over media funding, the risk indicators for media pluralism in Brazil are extremely high. Looking back with this perspective, the so-called disoriented audience already existed in Brazil, in search of greater diversity in news consumption.
It is in this context that the far-right media ecosystem in Brazil learned from the propaganda tactics deployed by former U.S. President Donald Trump’s 2016 campaign and supporters, including “junk news” websites, social media bots and trolls, and click-farms fueling disinformation at scale. In these terms, information and propaganda also come together as a political ideological tool, as they constitute an element of opinion formation and perception both locally, nationally, and internationally. Although there is a global far-right coordination and common infrastructure that impacts the information ecosystem worldwide, research developed in both Brazil and the U.S. has shown that the decline of local journalism also plays a critical role in eroding audiences’ trust in mainstream news.
A digital public sphere that became a propaganda marketplace
The internet ecosystem has been turned into a propaganda marketplace by the major tech platforms, which incentivize a variety of behaviors by groups and individuals that degrade the public discourse. Since 2016, scholars have identified critical dynamics impacting social cohesion and the digital public sphere. Books like Network Propaganda (2018) by Yochai Benkler, Robert Faris and Hal Roberts and Meme Wars (2022), by Joan Donovan, Emily Dreyfuss, and Brian Friedberg; the works of Alicia Wanless and Michael Berk on the participatory propaganda model; and countless papers by scholars such as Renée DiResta, Kate Starbird, Renata Gomes, Raquel Recuero, and Marie Santini are key to comprehending the participatory and immersive nature of the networked propaganda model transforming both culture and politics today. Propaganda is information weaponized by malicious actors who seek to manipulate the public by deploying persuasion techniques developed to interfere with and take advantage of society’s values, beliefs and behaviors.
Building from the long tradition of internet and media studies, I call the process of turning all aspects of our social interactions digitally traceable into a source of profitability through the logistics of recursive computation the algorithmization of life. Big tech platforms have spent a considerable amount of their resources creating the affordances that make the conditions of data extraction at social networking sites or in messaging apps work like a marketplace. Overall, the digital public sphere is now a multi-platform marketplace building in its contingency of immersive and participatory design.
From broadcast to digital media, the information crisis in Brazil connects the past to the present. Social media platform growth in Brazil is itself a history intertwined with the country’s own story of digital transformation and internet access. It was only in 2012 that Facebook.com (Facebook’s name at the time) ousted Orkut (Google’s biggest social media enterprise in Brazil for quite some time) as the country’s top social networking site . In 2012, Twitter had roughly 12.5 million visitors in Brazil, YouTube had only started to invest in Brazilian influencers, and TikTok and Kwai were still years away. Although internet access has grown disproportionately in Brazil’s urban and rural areas, this context reflects advances in digital infrastructure, broadband internet, and mostly smartphone growing consumers.
The labor force behind the propaganda marketplace
Artificial intelligence systems operating on social media platforms embed the modeling, prediction, and automation features in recommendation systems that consolidate current affordances to maximize revenue for their marketplaces. In the Brazilian political and cultural context, platform work has become the main source of income for many influencers over the years, and for some individuals and families, it’s the only one. Often people think of these workers only as Uber or Lyft drivers, but click-farms at social media platforms are also made of real humans putting their energy into tasks such as clicking, commenting, inviting and sharing content on social media networking sites.
The Brazil Fairwork 2021 report revealed the exploitation imposed by big tech and micro-work platforms on workers, from payment below the minimum wage to long working hours, violence in the workplace, health issues, and arbitrary deactivation. In the context of high rates of unemployment, companies in this economy promote informal work as a solution and subordinate workers who are self-employed, expanding informal work. Influencers, politicians and PR agencies realize this cycle depends on a robust engagement machine, and that’s where the automation features of the platform infrastructure affordances come in. If someone wants popularity, influence, and media impact, “real” followers or engagement is not a problem when it’s possible to just buy these services from click farm providers. Overall, the exploitation of a precarious labor force to satisfy big tech machine learning systems harms marginalized communities and individuals in Brazil, and has contributed to the growth of the far-right propaganda and disinformation industry.
Ignoring the writing on the wall
Unfortunately, before the 2022 Brazilian election, tech companies ignored warnings discussed at successive meetings and in numerous public reports about the dangers of the massive distribution of false content about the electoral system and incitation of violence against political figures.
In the runup to the election, the Brazilian Electoral Supreme Court (TSE) engaged with researchers, journalists, technologists, policymakers, educators, and civil society to formulate new policies for the General Election. Agreements were signed between TSE and popular social media platforms such as Twitter, TikTok, Facebook, WhatsApp, Google, Instagram, YouTube and Kwai. Reports addressed how Meta’s targeted advertising and marketing tools were used to amplify disinformation about Brazil’s voting system, and the YouTube recommendation algorithm’s bias towards far-right content. Mozilla’s Youtube Regrets Report also crowdsourced an investigation by building a browser extension, allowing people to donate data about the YouTube videos they regret watching on the platform. Results indicated that YouTube recommendations violate their own policy, and that non-English speakers are hit the hardest.
During and after the election, Brazil’s Digital Rights Coalition (CDR), a network of entities that brings together more than 50 academic and civil society organizations in defense of digital rights focusing on access, freedom of expression, protection of personal data and privacy online prepared a technical document calling out the role of social media platforms to protect election integrity..
The failure of big tech’s governance and policy mechanisms to to deploy basic interventions and prevent disinformation and hate-speech at scale are a springboard for anti-democratic actions. Evidence from Kenya and Brazil shows that systemic effects of big tech’s shoddy operations disproportionately impacts countries in the Global South because of power asymmetries. Companies such as Meta, Twitter, and TikTok, that reach billions globally, are failing in their duty to equally serve consumers around the world. A 2022 press release from the European Commission highlighted concerns about Meta’s decision to tie the social networking site (Facebook) to its online ads service (Facebook Marketplace). The trading conditions of ads-related data may be the future of platforms as they increasingly push boundaries of antitrust rules. Netlab, a research lab located at the Brazilian Federal University of Rio de Janeiro, launched a report based on open source methodologies highlighting which disinformation publishers profit from Google’s ad systems.
Overall, it is clear that big tech’s business model is contributing to a global information crisis, harming marginalized communities, and further obscuring issues that were already at risk of erasure in the information ecosystem. Social media platforms are part of the problem, and it is necessary to continue investigating the computational paradigms anchored in colonial expansion and other predatory practices at scale .
The fundamental issue is even more challenging because the for profit news business model is broken in Brazil, just as it is elsewhere. These circumstances demand a radical transformation of the political and economical approach to the public infrastructures that enable pluralism and access to information. By defining what “counts” as disinformation or “post-truth”, the parameters of ethics and interpretation of history determined by them are revealed. Thus, the categorization of disinformation itself maintains a commercial model of media that strategically concentrates on the same decision-making institutions and actors indicating what counts as information or not — perpetuating existing inequalities and disregarding certain information as unrelated to the public good. As long as the debate on these issues is championed by a handful of influential corporate media and big tech players, it will continue to reinforce the racial, economic and cultural asymmetries at the root of the information crisis.
Global concerns and hyper-local pathways for media and tech reparations
A critical approach to the categorization of disinformation must attack the fundamental problems of the Brazilian communication media model and align with a necessary debate on race that is not restricted to diversity in the newsroom and tech companies. Issues such as editorialization, moderation and accountability must be treated in terms of shared responsibility over information as a human right, consistent with Brazil’s 1988 constitution. In times of doubt, uncertainty and distrust of democratic and scientific institutions, the problem of disinformation demands a reconfiguration of hegemonic spaces that define discourses and the production of knowledge, leveraging perspectives and voices erased or deliberately made visible through harmful stereotypes. Critical disinformation studies propose grounding an approach in history, society, culture, and politics, positioning race, gender, class, and territorial analysis to understand how these signifiers shape dynamics of disinformation, all while investigating how institutional power and economic, social, cultural, and technological structures shape the information crisis, and reflecting on how grassroots communication practices, mobilization through traditional media such as radio, hacking technologies, and various local solutions can be strengthened to serve those impacted by the violence that often accompanies disinformation.
Over the last 20 years, data colonialism led by big tech and the ideology of progress in Silicon Valley has laid the ground for the violations of individual and collective rights at ecological and societal levels. In 2022, an inability to confront this colonial logic embedded in tech platforms kept the Brazilian civil society representatives seated at meetings with trust & safety, integrity, and government relations’ big tech teams constantly overwhelmed by the blurriness of their own terms of negotiation. This set of circumstances has disproportionately impacted countries in the Global South, following the pervasive harms of globalization as described by the Brazilian intellectual Milton Santos. The social, geographic, and cultural practices matter in context considering language and regulatory specificity by regions. The solution will not come top-down from companies in Silicon Valley through negotiation terms dictated to their teams located in Brazil. Simply put: we can not pressure for regulations or reparations in regards to the information crisis in Brazil without access to meaningful data, taxonomization of the problem, and labeling to support independent auditing work.
The broadcast and digital frameworks at the core of the information crisis are closely linked by the structural roots configuring the concentration of power and discourses. The category of disinformation– reduced to what some call the “crisis of expert systems”– continues to mark the authority of a few while marginalizing many communities. When W. E. B. Du Bois (1935), in a book chapter entitled “The Propaganda of History”, questioned the ways in which the reconstruction of the United States was being studied and taught at the time, he queried what the standards of ethics in research and interpretation for reading history would be. According to Du Bois, history was used not in a scientific way, but as propaganda to inflate the nationalist ego. Thus, by hiding the facts and making the history of the struggle of Black people for freedom and democracy invisible, history generated a false sense of accomplishment for white people, who continued to tell a certain story. His criticism is valid for the contemporary debate on disinformation, as it is not just a crisis of experts or a “post-truth” phenomenon – since obviously, we did not live in the “paradise of truth” before Donald Trump or Jair Bolsonaro. In the case of the current information crisis, broadcast, digital, legacy media and social media platforms continue to pressure and guide the parameters defining what counts as disinformation, framing the information crisis as solely a gross scandal of the far-right and Bolsonarism in order to keep certain conversations out of the broken commercial system.
In 2023, the battle to protect democracy and address the climate emergency harming disproportionately racialized communities must also deploy principles of equity, recognition and representation of plural interests in media and technology. The failed attack on Brazil’s democracy on January 8th highlights that, more than ever, these concerns are not only urgent, but also demand transnational collaboration between civil society and grassroots movements fighting for media and tech reparations.
The pathways to building a stronger democracy are hyperlocal due to the shared need to reclaim history in the context of capitalism and racial subjugation. Leveraging community based journalism practices grounded in the many ways that regular people communicate and connect in Brazil must be addressed so we don’t reproduce the same colonial paradigms regionally. Non-profit initiatives such as the Amazon Radio News Network and the Living Map of Amazonian Media show that combining local experience with collaborative tech infrastructures and content collaboration can combat disinformation and expand audience exposure to news that engenders trust and advances pluralism. The Working Group Fighting Disinformation in the Legal Amazon led by Intervozes is another important initiative calling out the need to demonetize the hyperpartisan online news ecosystem that promotes disinformation about territorial rights and perpetrates attacks against local environmental defenders and indigenous communities and leaders. Outside Brazil, transnational collaborations between researchers and projects in tech and media, such as Derechos Digitales, Fix Media Now by the Media and Democracy Project, Hope not Hate, Climate Action Against Disinformation, Conscious Advertising Network, Design Justice Network, Distributed AI Research Institute can also pave the way for a global alliance rooted localized justice principles and strategies.
It is necessary to advance such initiatives if we are to address the true underlying problems in Brazil’s information ecosystem, and produce a more meaningful and inclusive public dialogue that advances democracy. In the wake of the January 8 attack, it is crucial that this work begins in earnest.
– – –
Regattieri would like to thank Becca Ricks (Mozilla Foundation), Koliwe Majama (Mozilla Foundation) and Rafael Grohmann (Toronto University) for their generous review of and contributions to this essay.
Lori Regattieri is a social-environmental-climate justice tech and movement builder, and is presently Senior Fellow in Trustworthy AI at the Mozilla Foundation. As an activist and communications advisor, she has more than 15 years dedicated to campaigns, mobilization and collective action supporting grassroots movements in Brazil. She’s also a research associate at Netlab (UFRJ), a member of the Network of Latin American Studies of Surveillance, Technology and Society (LAVITS), Design Justice Network (Allied Media) and the VOX-Pol Network of Excellence (NoE) focused on researching Violent Online Political Extremism. Lori holds a B.A. in Social Work and MA in Communication and Territoriality from the University of Espirito Santo (UFES/Brasil). She spent a year abroad at the Digital Humanities Program, University of Alberta, Canada. In 2021, she completed her PhD in Communication and Culture from the University of Rio de Janeiro (UFRJ) with the thesis “Algorithmization of life”. Over the years her research has focused on science and technology studies (STS), cybernetics, media ecology, propaganda and disinformation, reparation epistemologies and decolonial approaches to increase justice and reduce bias in AI. She is the co-editor of the feminist magazine DR. More info at http://eco-midia.com and at http://netlab.eco.ufrj.br.