Society | Promoting education for digital engagement and media literacy, along with strengthening ties between journalists, scientists, and the public, are among the recommended strategies to combat misinformation
*By: Cristiane Miglioranza
*First published August 15, 2024
*Photo: Freepik
Five out of ten Brazilians frequently come across news that seems fake. In addition, 36.5% admit they have shared false information with friends, relatives, or online, regardless of whether they suspected its veracity. The data come from the 2023 edition of the survey Percepção Pública da C&T no Brasil (Public Perception of S&T in Brazil), which interviewed 1,931 people aged 16 or over from the five regions of the country.
Developed with the aim of understanding the population’s views, interests, and level of knowledge regarding Science and Technology, the survey has been conducted since 2015 by the Center for Management and Strategic Studies (CGEE), based on a historical series that began in 1987. CGEE is a social organization supervised by the Ministry of Science, Technology, Innovations and Communications (MCTIC). In the study mentioned, it received support from the National Institute of Science and Technology – Public Communication of Science and Technology (INCT-CPCT) and from the Brazilian Society for the Advancement of Science (SBPC).
The CGEE survey points out that the COVID-19 pandemic marked a turning point in Brazilians’ information consumption behavior. Until 2015, television was the most used medium of information. After the pandemic, social media, messaging apps, and digital platforms, which were already gaining ground in preference, became the main sources for seeking news about Science and Technology, used by 39.8% of respondents. Among the most popular are Instagram, Facebook, YouTube, and WhatsApp, cited by 73% of respondents.
According to the data, fact-checking is an uncommon practice among the public and is strongly related to education level. This practice is less frequent among people who live in rural areas and also varies with religion and region of residence.
The phenomenon of disinformation among Brazilians was also the subject of a study conducted by the Locomotiva Institute, which focuses on developing strategies for companies, brands, and institutions interested in engaging with their audiences. The Locomotiva Institute study shows that 88% of the Brazilian population over 18 years old has already believed in some type of disinformation related to product sales, electoral proposals, public policies, political scandals, and fake content about the economy, the public security, and the prison system.
According to respondents, the greatest risks posed by disinformation are electing bad politicians, damaging someone’s reputation, creating fear concerning security, and harming health. The study reports that 35% of those who realized they had been deceived felt naive, 31% felt angry, and 22% felt ashamed. Another interesting finding concerns content dissemination: 24% of respondents said they had been accused of spreading fake news by people who disagreed with them.
But after all, what is disinformation?
The concept of disinformation encompasses a range of practices that go beyond the production and dissemination of fake news. Disinformation refers to the use of communication and information techniques to mislead or create a false perception of reality by suppressing or concealing information, minimizing its importance, or changing its meaning, with the goal of influencing public opinion and protecting private interests.
The Brazilian Academy of Sciences (ABC) understands disinformation as a phenomenon that transcends geographical frontiers and knowledge areas. It affects people’s ability to make informed decisions and impacts their trust in governmental scientific institutions. Since 2023, the ABC has had a specific working group to study the phenomenon, and the first report about the subject was coordinated by researchers Glaucius Oliva (University of São Paulo – USP), Ruben Oliven (Federal University of Rio Grande do Sul – UFRGS), Roberto Lent (Federal University of Rio de Janeiro – UFRJ), Virgilio Almeida (Federal University of Minas Gerais – UFMG), and Thaiane Oliveira (Fluminense Federal University – UFF).
A global phenomenon impacted by artificial intelligence
In June 2024, the Organisation for Economic Co-operation and Development (OECD) published the Truth Quest Survey, which measures people’s ability in 21 countries to identify false and misleading online content. Brazil had the worst score, 54%, while Finland had the best overall performance, with 66% accuracy in assessing content veracity. The study shows that Latin American countries need to improve their social media literacy. Moreover, more than 85% of people in Latin America obtain news from social media, in contrast to less than 60% in Germany, Japan, and the United Kingdom, for example.
The data show that the ability to assess the veracity of information was higher in countries where people consume less news via social media. In addition, the veracity of AI-generated content was identified more easily (68% accuracy) than content produced by humans. According to Ana Regina Rêgo, a professor at the Federal University of Piauí (UFPI), addressing disinformation must be combined with literacy in the use of Artificial Intelligence (AI). “Nowadays, people encounter an increasing amount of AI-generated content,” emphasizes Rêgo, coordinator of the National Network to Combat Disinformation (RNCD).
Formed in 2019 based on Rêgo’s research, the RNCD is a collective of academic, professional, and activist initiatives that develop and present solutions to address rumors, biased narratives, and falsehoods. The researcher draws attention to the issue of ethics. In an essay published in January 2024 in the journal Organicom, from the School of Communications and Arts of the University of São Paulo (ECA/USP), Rêgo reflects on digital platforms and the phenomenon of disinformation. “To a large extent, AIs are fed by agencies based in the Southern Hemisphere that pay low wages to the professionals who feed them. Another concerning factor is the transfer of prejudiced biases to generative language models, which reverberate machismo, misogyny, fascism, homophobia, xenophobia, among other aspects that are capable of generating both texts containing disinformation and hate speech,” she emphasizes.
Rêgo analyzes that this transfer of prejudices rooted in the social context is one of the main reasons for both distrust in AI and concern about its potential to generate discriminatory discourse.
“The way is not to oppose technologies that arrive at an ever-increasing speed, but to seek viable alternatives for dialogue and negotiation with such techno-market structures through regulation and development of alternative and national digital platforms that can stand up to the big techs, which position themselves in the global geopolitical scene as parastates.”
Ana Regina Rêgo
Many social networks, such as Instagram, allow content posted by a user to be identified as generated by Artificial Intelligence. However, this initiative must come from the person who posts it, because there is no control over it. “Thinking about the development and possible uses of Artificial Intelligence in the communicative field requires the adoption of ethical principles that represent social respect within a techno-marketing environment that is at the service of society, rather than the other way around,” Rêgo recommends.
Science versus disinformation
The document published by the ABC in June 2024 defines scientific disinformation as the dissemination of false, misleading, or inaccurate information about scientific issues, frequently related to health, environment, or technology. This type of content is crafted to appeal to emotions, personal beliefs, and prejudices, rather than being grounded in solid evidence.
Scientific disinformation is identified as a practice that often exploits vulnerable points in society, creating a conducive environment for the dissemination of false information, conspiracy theories, and fake sciences. These are a subset of scientific disinformation that seeks to emulate the credibility of valid research, but, in reality, does not follow the principles of scientific methodology. It is conducted with bias or misleading intentions. One recent example of fake science is the recommendation of hydroxychloroquine and ozone therapy for the early treatment of COVID-19.
Rêgo highlights that, according to research conducted by the Massachusetts Institute of Technology (MIT), disinformation has 70% more potential to go viral than accurate information. “Organically, disinformation reaches one hundred thousand people, while accurate information reaches up to one thousand. This happens because the business model of the platforms allows for a boost, that is, it recommends this type of content through algorithms, based on other content that a person consumes,” she explains.
According to the glossary of the Getúlio Vargas Foundation (FGV), ‘organic reach’ is an expression used in digital marketing to describe the number of people who view content without paying for it, that is, without the need for the creator to invest in advertising. To understand how organic reach works, it is necessary to know how search engine algorithms work, such as Google’s. These algorithms are responsible for classifying and displaying the most relevant results of users’ searches. When a person conducts a search, the algorithm analyzes several factors, such as the quality of the content, the relevance to the search, and the authority of the website. Based on these criteria, it determines which results are displayed.
“Algorithms know us better than we know ourselves. So, a digital platform not only exposes us to a certain type of content but also recommends and monetizes it. As it appeals to emotions and has a confirmation bias, disinformation is a highly viral and shareable type of content. We saw this during the pandemic.”
Ana Regina Rêgo
The RNCD coordinator considers the belief that artificial intelligence will solve humankind’s problems to be a fallacy. “I believe it is actually exacerbating existing issues, including those related to the production and dissemination of knowledge. In Brazil, there are more than 30 million people without access to the internet and more than 86 million people with poor internet access. If someone cannot afford a basic internet plan or lives on a minimum wage, how could they check something that they are unaware of?” Rêgo asks.
Rêgo points out that technologies are not neutral and, therefore, need regulation. “The RNCD has a large group of civil society and academic collectives that monitor and act for the regulation of platforms and Artificial Intelligence. Last year, we worked a lot on Bill No. 2630/2020 [known by its supporters as the Fake News Bill and by its opponents as the Censorship Bill], which the Speaker of the Chamber of Deputies, Arthur Lira, shelved, and we are trying to bring it back. It is difficult now due to the elections. There is also Bill No. 2338/2023, which provides for the use of Artificial Intelligence. So, we not only have researchers, but also activists.”
The ABC emphasizes emotional appeal as a distinguishing feature of disinformation messages. It highlights that AI-based techniques for automated fact-checking can include emotion recognition in the textual or visual components of disinformation. According to the document Desafios e estratégias na luta contra a desinformação científica (Challenges and strategies in the fight against scientific disinformation. Free translation), “AI can perform tasks, such as determining if the text expresses a specific emotion (anger, fear, sadness, or joy); identifying the polarity of the text (whether it is positive, negative, or neutral) through a task commonly called sentiment analysis; establishing the position of the text regarding some aspect (such as the polarity towards a certain subject or whether the text presents a favorable or opposing opinion about this subject); identifying whether the text presents emotional appeal, among other possibilities.”
Regarding visual information, the AI used to combat disinformation aims to automatically identify whether the image in the news presents an emotional appeal or whether a person’s image shows some kind of emotion. “However, it is necessary to consider that identifying emotional aspects in isolation, based on textual and visual information, is a highly error-prone task. The perception and understanding of an individual’s emotional cues demands time of interaction, knowledge of the person, history, personality, and other characteristics that are not present in texts or images, such as tone of voice, gestures, and body cues,” the ABC highlights.
The ABC emphasizes that AI methods must be used cautiously and rigorously before their responses are considered true. “When used by malicious individuals, these methods can be directed to generate false information on a large scale, including characteristics often found in disinformation […]. Therefore, it is essential to develop strategies that prevent the indiscriminate dissemination of AI-generated data and promote the development of tools for identifying and tagging machine-generated content,” the document states.
Development of information and communication competencies
Besides regulation and fact-checking, the Brazilian Academy of Sciences recommends media and information literacy and education as a way to address scientific disinformation. Within this context, Jussara Borges de Lima, a professor in the Graduate Program in Information Science (PPGCIN/UFRGS) and at the School of Library Science and Communication (FABICO/UFRGS), coordinates the project “Educação para a informação: uma proposta de enfrentamento à desinformação” (“Education for information: a proposal for addressing disinformation”).
Head of the Research Group on Behavior and Information and Communication Competencies (InfoCom), Lima emphasizes that disinformation is one of the most acute problems nowadays. According to her, healthy and ethical appropriation of information is necessary in a society of prosumers. The concept of prosumer was developed in the context of marketing to handle changes caused by social networks and digital platforms, and it gathers content producer and consumer in the same person.
“We are not just content consumers anymore. We are producing content without realizing it. When you click or like something, you are producing content, even if it is for a company that paid for the algorithm. This is the behavior of a consumer who not only wants to consume but also wants to give something in return. As researcher André Lemos says, people not only want to watch the show. They also want to participate; they want to put on the show.”
Jussara Borges de Lima
Lima emphasizes that there are a number of studies aimed at understanding the phenomenon of fake news and fact-checking mechanisms, but there are few that address disinformation. The researcher has been working with the concept of information and communication competencies for 15 years, seeking to foster a more critical approach to information consumption. “That is, not just to teach people to look for reliable sources, but also to reflect on it: why is this source better? Why do I need to evaluate information? Why is veracity important?”
As a professor at FABICO, she observes the difficulty students face in producing more authorial and creative content. “Most of the time, students come from backgrounds in which school assignments were based on ‘copy and paste.’ When I propose a question for them to solve in a creative way, they face difficulties such as ‘Where do I find the answer?’ I always tell them they have to develop and create that answer; it does not exist yet.”
The project Lima coordinates is based on the premise of research, teaching, outreach, and innovation. “What can we do at the university? How can we promote education for information? How can we transform this into a public policy that reaches more people?” From these questions, her research group works in the field of outreach with a course aimed at librarians, archivists, communicators, and students, with a group beginning in October 2024.
The training takes place every two years, and spots in previous editions filled up quickly, indicating pent-up demand. “Besides issues regarding the content of information, classes also address people’s relationship with the content, with the meaning. They work on collaboration and even more, let’s say, targeted aspects. For example, how to deal with hate speech and how to practice nonviolent communication. The course focuses on these competencies in the field of communication,” she explains.
The professor points out that disinformation has always existed and has grown in scale and impact with the internet. “We already had elections defined by disinformation. In the economic field, there are studies demonstrating that Brexit [the withdrawal of the United Kingdom from the European Union] was a decision widely supported by disinformation. In the health field, people died as a result of disinformation about COVID-19. This happened at the macro level,” she says. “At the micro level, I think all of us have been part of those family WhatsApp groups in which relationships were broken because lies, deception, defamation, and calumny were spread. In the end, it caused a series of conflicts that impacted relationships and lives,” she adds.
At the end of each edition of the outreach course, evaluations were conducted with the participants. The feedback demonstrates changes not only in interactions with content and people, but also regarding their social role in this context. “Professionals and students who participate in the outreach course start to realize that they can also be educators, and that this is important in the environment they live, work, and act,” Lima says.
“Our spots fill up in two hours – demand is high. It means society is realizing the need, and we cannot meet it. We need a public policy that starts attending children at Elementary School and reaches all the public. Older adults also need training in this area because they are major victims of disinformation.”
Jussara Borges de Lima
In this regard, the researcher highlights the Brazilian Media Education Strategy (EBEM) as an interesting federal public policy initiative, despite its name. “Education is not aimed at media, but at the content and the relationships mediated through them. But all paths are important, and we cannot discard them: whether it is research, outreach, science popularization, or public policy.”
Lima believes in outreach activities as a bond between scientists and the community. “It is not only important for offering solutions but also for us to learn with the community. This is because people are also finding solutions within their reality and context,” she emphasizes. Concerning the idea of science popularization, she highlights the role of those who translate scientific language for the general public and the need for more investment and educational initiatives accessible to the general community.
Producing information is not enough
Fábio Pereira, a researcher at the University of Brasília (UnB) and at the Université Laval in Quebec, Canada, points out the gap between science and society. According to him, journalism needs to rethink its concept of audiences not only to combat disinformation, but also to reconnect people with scientific knowledge and reestablish their trust in it. Based on his research, he observes that discourses tending to say science is in crisis are not confirmed in reality. “We are beginning to realize that science is indeed important and it plays a part in people’s daily lives and decision-making,” he says.
Pereira, a specialist in science journalism, emphasizes that the key issue is content. This means that people truly want to know about science that is more connected to their daily lives. “How do you communicate science effectively? How do you explain in simple terms what scientists are doing? How do you help people know the structure of the scientific field? We also need to make an effort to understand what the audience is looking for in terms of scientific information so that journalism can be a source again,” he reflects.
Reconnecting with the audience presupposes that it is diverse and has different levels of engagement with the topic. The researcher explains that there are at least three types of audiences in the context of science popularization: “There is a specific audience, which is formed by people who are really fond of science and come from a socializing process that allows that. We use the term ‘amateur’ not to mean someone who casually engages in an activity, but in the sense of a person who loves that content. In fact, this is the audience that identifies itself as belonging ‘to science.’’’
There is also the audience that searches for science because it is useful in everyday life. That is the case of people who are interested in information about health, for example. The third type of audience gets informed during moments of crisis, such as the pandemic or global warming. “This is an audience that needs to know what is going on, but has never been interested in science before,” Pereira says.
The researcher highlights that people who have a more peripheral relationship to scientific information and science journalism are the most susceptible to disinformation traps. According to him, both academia and journalists need to move together.
“Indeed, we need to take action to bring science back to society. When communicating information, we need to think not only about the value of a discovery within the scientific field, but also understand whether it affects people’s lives and what games and social dynamics are involved in the production of this knowledge,” he says.
“Scientists, especially, have a very authoritative view of science: ‘I have discovered it, so it is important.’ Knowledge production is also part of social dynamics. The researcher’s choice of the topic and of what receives funding is absolutely social. This kind of social dynamic is hidden when the results are shared, making them appear as the individual merit of a team.”
Fábio Pereira
According to Pereira, the restoration of science production conditions and the exposition of the effects it causes in society humanize the process. “It is a way of saying that we, scientists, are not an isolated group in society. We are workers, and we feel like workers, just like you.”
The researcher observes that people who deny science or are drawn to conspiracy theories make up a smaller minority than it might seem. “A study conducted by a colleague from the Université Laval in partnership with a Canadian radio station analyzed audience comments on Facebook during the pandemic. He discovered an interesting behavior: the users themselves corrected information that showed denialist or conspiratorial bias. Many of the people commenting there were working within a logic of seeking consensus,” he reports. “This shows that we, journalists, and to some extent, scientists, have a stereotyped and slightly negative view of people’s ability to understand what is going on,” he warns.
The second point the research shows is that a large part of the audience wants to be heard, to participate in the conversation with the journalist and the scientist, and to connect with these information producers. “We, journalists, were trained to produce information. We believe that the credibility crisis is due to a lack of information and conclude that people are not reading. There is too much information, and nobody can handle it all. So, we need to anticipate moments of interaction with the audience in those spaces of connection as a way to try to listen to them,” he concludes.
Translated into English by Cristina Gueths de Aragão, undergraduate student enrolled in the course “Supervised Translation Training I (English)” of the Undergraduate Program in Language and Literature, under the supervision and translation revision of Professor Rozane Rebechi (P.h.D.) – IL/UFRGS.
:: Read in portuguese