#Hashtag: Platforms don’t want to commit profit to fight fake news, says researcher

Eight months before the presidential election, digital platforms have said little about how they will face the challenges of disinformation during the election period. In the agreements signed with the TSE (Superior Electoral Court) this week, many of the measures presented are not specific to the Brazilian election and fall short of the policies adopted in the US.

For the anthropologist and researcher David Nemer, who integrates the Berkman Klein Center for Internet and Society at Harvard University (USA), the lack of commitment of companies can be explained by economic interests. “They don’t really want to act in a way that reduces engagement, as fake news generates engagement, and engagement is the way they monetize. They don’t want to compromise profit with measures that can reduce the effect of disinformation.”

According to him, even after the campaign that led to the invasion of the Capitol on January 6, 2021, platforms do not seem very concerned about the political risks of fake news in Brazil. “They feel very safe to do these things,” he says.

Nemer proposes that social networks adopt measures to identify the so-called “disinformation hubs”, a set of accounts that orchestrate the campaigns of lies on the networks. According to him, these profiles are usually responsible for taking fake news from the “periphery” to the “center” of public debate on the internet.

He gives as an example the message that has circulated on the networks in recent days about Bolsonaro having avoided World War 3 by traveling to Russia. “One account was leading this campaign. You could see how one person, a large account, was enough. By being able to identify and remove this account, you mitigate the effects of misinformation”, he defends.

The researcher also describes how the Bolsonar disinformation apparatus has transformed over the years, from a focus on WhatsApp in 2018 to an escape to Telegram and YouTube in 2022.

Digital platforms announced few specific measures for the Brazilian election in agreements with the TSE on fake news, especially in comparison to the scenario in the US. What justifies this attitude?

The measures that have been proposed are very inefficient and show the real commitment of the platforms in relation to disinformation and fake news.

That is, they don’t really want to act in a way that decreases engagement, since fake news generates engagement, and engagement is the way they monetize. They don’t want to compromise profit with measures that can reduce the effect of disinformation.

The proposals are “slap in the glove” in the face of people who are studying, criticizing, even to help platforms understand how to combat disinformation. It’s frustrating.

Aren’t the political effects of disinformation harmful to companies?

They can be, yes. But so far, in Brazil, it seems that they are not worrying much about it. Until then, they feel very safe to take these attitudes.

With the exception of Twitter, none of the companies responded as to how they will react in case of contestation of results and incitement to violence. Can they be held accountable if these threats materialize?

They can, yes, because the Civil Rights Framework for the Internet, for example article 19, makes it possible to legally request that the platform remove content. If the platform does not withdraw, it can be held responsible for that content.

Even so, in Brazil, with the Marco Civil, it is still understood that platforms, in principle, are just a means. They are not legally responsible for the content. But after a court order to remove unfulfilled content, there comes the responsibility. But a priorithey are not responsible for the contents, only the user.

don’t you think would be better, after episodes such as the invasion of the Capitol last year, we have measures that most prevent rather than cure?

Certainly. It is much better to be proactive than reactive. And in Brazil there is a very strong potential for what happened here, in the United States, to occur.

o Barroso [presidente do TSE, Luís Roberto Barroso] is trying to do this. When he engages in this debate with social networks, the TSE wants to be proactive, he wants to understand how to combat disinformation. It is only up to us to be hopeful, but in the face of these answers [das plataformas] there’s not much to be optimistic about.

Bolsonaro stopped for now, but on WhatsApp and on the Bolsonarista Telegram every now and then this issue of the electronic voting machines being hacked reappears. They say they won’t accept [o resultado das eleições] if there is no printed vote. A conversation that we saw more at the end of last year, but it is something that Bolsonaro will resume. He’s a little calmer about it because the TSE pulled the reins and spoke firmly.

So you believe that platforms really have the ability to contain disinformation on networks.

Have. It is not possible to contain 100%. It is impossible to scan the net, identify and remove everything. But you can identify these “hubs”, the accounts of disinformation. That’s where most of the disinformation and orchestration comes from. They are the ones who bring disinformation to the center of public debate. Once removed, these themes return to the periphery of the debate and do not have the damage they would have if they were part of the central debate.

It is possible to end the consequences of misinformation. Platforms can use actions with pedagogical attitudes. And as a warning: take someone as an example and withdraw.

How should Bolsonaro and Bolsonarist groups act if they don’t win the election?

That is the great mystery and the great fear. Today, the approval of the Bolsonaro government is minuscule. Far from being a majority, but enough to take people to the streets. The anti-democratic protests demonstrated this. They are not a political force to the point of being a national movement, but enough to bring people together to do damage. They have the potential to engage in totally undemocratic attitudes.

In these elections, social anxiety will be much greater than in 2018, as Bolsonaro delivers the country to an unprecedented crisis. His campaign will have to recreate a reality to convince his base to vote for him.

From 2018 onwards, do you see any difference in the behavior of groups that support the president on the internet?

In 2018, people were being paid to develop fake news for WhatsApp. Today, with the advancement of the fake news inquiry and with the CPMI in Congress, this dynamic has changed a lot. The money that financed disinformation on WhatsApp no ​​longer exists, as those who financed it no longer want this exposure.

With the withdrawal of this funding, those who produce disinformation sought other ways to monetize. Even before being there for a political ideology, people are there for the money.

In 2019 and 2020, they started to enter the business of websites, such as Jornal da Cidade Online, which brought WhatsApp traffic. These sites monetized on top of Google Adsense ads. With the advancement of Sleeping Giants, for example, and campaigns against these channels, advertisers began pulling ads from these platforms.

So the next step was YouTube. Today, within WhatsApp and the Bolsonarista Telegram, what circulates the most are links to YouTube videos, because there these disinformers manage to monetize. YouTube pays them for every click, every visit.

If YouTube demonetizes these groups, the disinformation on the platform ends. Making disinformation is work, it takes time, resources and nobody does it for free. YouTube seems to be their last hope.

Telegram has ignored the decision of the STF and does not return the contacts of the authorities. Should the app be blocked in Brazil?

It is a situation that is not good for anyone. You have an app that is above any law in the country and that will be a strong instrument of disinformation [na eleição]. But, on the other hand, it is also a platform that is in 53% of all cell phones, which is practically a public utility, where people get information. It is an extremely complex and extremely important application.

So letting it run wild is not going to be good for our elections, and neither is banning. I think there has to be conversation [com as autoridades]. But if one side refuses to talk and sees itself above our Constitution, then it becomes very difficult to criticize Barroso or any authority. Only that kind of attitude will remain [o banimento].

You are creating a robot that reports hate speech. Can you talk a little about him?

The way these platforms fight fake news is far behind. It is possible to find groups on Facebook that celebrate Nazism, for example. With the bot I’m testing, for example, the goal is to know what kind of speech Twitter takes most seriously to take down. So far, what I’ve come to understand is that homophobic speech seems to be less tolerable, because the accounts that were suspended in that experiment were accounts that engaged in homophobic speech.

X-RAY | DAVID NEMER, 38, VICÓRIA (ES)

He is an anthropologist and researcher at the Berkman Klein Center for Internet and Society at Harvard University, USA. He is also an Associate Professor in the Department of Media Studies and Latin American Studies at the University of Virginia, USA. He has just released the book “Technology of the Oppressed-Inequality and the Digital Mundane in the Favelas of Brazil”, by Milfontes.