Advertisement
News

WhatsApp is managing its fake news crisis in Brazil from an office in Silicon Valley

“It gives us a sense of how important we are.”

by David Gilbert
Oct 26 2018, 6:14pm

Last Wednesday, after facing mounting criticism over WhatsApp's role in the spread of fake news during Brazil's high-stakes presidential election, the company's new CEO Chris Daniels wrote an op-ed to Brazilians, promising to do better.

“We have a responsibility to amplify the good and mitigate the bad," Daniels wrote in Folha, one of Brazil’s largest newspapers.

Five days later, when employees from Twitter, Facebook, and Google met in the Superior Electoral Court in Brasilia to discuss the impact fake news was having on the country’s election, WhatsApp didn’t even show up.

Rather, Keyla Maggessy, WhatsApp’s senior manager working on law enforcement and safety, video-conferenced in from her office in Silicon Valley. The company's failure to attend in person was compounded by what fact-checkers present in the meeting described as Maggessy’s lack of knowledge on the subject.

“Every question we put to her, she either didn't know how to answer or she was going to ask somebody else,” said Christina Tardáguila, director of fact-checking group Agência Lupa, who attended the meeting.

“It gives us a sense of how important we are.”

Tardáguila was one of two people present at Monday’s meeting in Brasilia who described Maggessy as underprepared. The other, Tai Nalon, director of Aos Fatos, Brazil’s oldest fact-checking network, said Maggessy wrongly claimed that WhatsApp had been working in partnership with her group.

“You may have a partnership with me, but I don't have a partnership with you,” Nalon said.

Maggessy did not respond to requests for comment, but WhatsApp strongly denied the idea that she was unprepared, saying she has worked closely with the Electoral Court in Brazil in the lead up to the first round of the election and after it. WhatsApp also said they were notified about the meeting just over 4 hours before it began, though they did apologize to Nalon for the partnership gaffe. A company spokesperson added: “We deeply respect the role of the electoral courts and have worked hard to explain our approach and the steps we are taking to support users in Brazil.”

But for fact-checkers in attendance that day, What’sApp’s lack of a physical presence on the eve of a presidential election was symbolic. Despite having 120 million users in Brazil, the company has no official employees in the country and those they do have monitoring fake news are all in Silicon Valley.

“It gives us a sense of how important we are,” said Tardáguila.

As Brazilians head to the polls on Sunday to decide between far-right strongman Jair Bolsonaro and left-wing Workers Party candidate Fernando Haddad, the encrypted messaging app has found itself thrust into the spotlight. And WhatsApp’s performance Monday is just the latest example that the company is not taking its role in the country seriously enough, fact-checkers and researchers in Brazil said.

Last Thursday, just 10 days before the vote, an explosive report by Folha de São Paulo claimed that powerful backers of Bolsonaro were providing illegal, undeclared campaign donations by spending millions of dollar to spread smears against his rival Haddad via WhatsApp.

Read: Brazilian politics is backsliding toward militarism, and the presidential election could make it worse

At the heart of Folha’s report was a Cambridge Analytica–like operation by a group of marketing companies who collected databases of WhatsApp users phone numbers that they would then spam with anti-Haddad fake news.

One of the popular fake news items currently being spread about Haddad on WhatsApp is that he equipped schools with baby bottles featuring a penis-shaped teat in an apparent bid to fight homophobia while he was mayor of São Paulo.

“This is what PT and Haddad are preaching to your kids,” a man declares in one video while showcasing the fake bottle.

Researchers say these are some of the latest examples of how WhatsApp has been weaponized by propagandists to deliver fake news and hate speech during an already divisive election season. They argue that the company is not taking the situation seriously enough and is unwilling to implement changes. But WhatsApp says many of the allegations being made have not been backed up by proof, and that if there were such a significant operation, their system would have detected it.

Given the encrypted nature of the network, investigating how widely abusive content is shared on the network is virtually impossible. While misinformation on Facebook is for the most part out in the open, on WhatsApp it is hidden among private chat groups, spreading unchecked among family and friends.

Absentee landlords

Presidential candidate Jair Bolsonaro reacts during a news conference at a campaign office in Rio de Janeiro, Brazil October 25, 2018. REUTERS/Ricardo Moraes
Presidential candidate Jair Bolsonaro reacts during a news conference at a campaign office in Rio de Janeiro, Brazil October 25, 2018. (REUTERS/Ricardo Moraes)

For years, experts have warned that the popularity of WhatsApp in Brazil made it an ideal target to be hijacked by those spreading fake news during an election. A recent poll revealed that 44 percent of people get their political news on the service, and its central role in Brazilians’ lives makes it a more trusted outlet than traditional media for many.

All the more reason the company should have a physical presence in the country, fact-checkers argue. Facebook, WhatsApp’s parent company, for example, has multiple offices in the country.

The fact that WhatsApp has no employees based in Brazil isn’t an unusual situation, said a company spokesperson. Even though WhatsApp is a global service, almost all of its staff are based at it the Menlo Park headquarters it shares with Facebook. Following a fake news scandal in India, where WhatsApp has been blamed for mob killings, the company announced it was putting a team in place on the subcontinent. But that’s an exception rather than the norm.

WhatsApp says it has people across engineering, legal and policy departments working on the Brazil issue since January, but they would not give a hard number.

Black box

Facebook and Twitter have ramped up their purge of fake news on their platforms, but their efforts have had unexpected consequences, experts said.

Tardáguila said that people had already begun to be “very cautious about what was being said on Facebook, on Twitter, and through links and articles on Google.” But the fake news clampdowns at Facebook and Twitter have accelerated the migration of misinformation to WhatsApp, she said.

“Of course, that sent dirty campaigns to WhatsApp,” she said.

On WhatsApp, Brazil’s fake news crisis got even trickier. “If we have an app that facilitates conversations with their relatives and communities, it is expected that what circulates on WhatsApp will be considered more relevant — even if it is fake news,” Nalon said.

Read: Brazil's populist candidate for president is getting a boost from an alt-right social network

The problem is essentially baked into the design. Put simply, WhatsApp is not a social network; it’s a messaging app, and its main draw is the end-to-end encryption it offers to its users. This feature is great for protecting users privacy but terrible when it comes to policing fake news, as even the company can’t say with any certainty what type of content is being shared on its network.

“Of course, that sent dirty campaigns to WhatsApp.”

Depending on what research you read, the results are wildly different.

One analysis published last week by researchers involved in tracking fake news on WhatsApp in Brazil concluded that 56 percent of the most popular memes shared on public chat groups were misleading. However, a survey conducted this week by market research company Ibope claimed that 75 percent of respondents had not received any negative content on WhatsApp from any presidential candidate. And of the 25 percent who did, only a quarter said the messages would impact how they vote.

Following the allegations published in Folha, WhatsApp sent cease-and-desist letters to the marketing companies alleged to be conducting the smear campaigns — as such practices are against its terms of service. The company says it has also blocked hundreds of thousands of abusive accounts — including Bolsonaro’s own son Flávio — in the run-up to the election.

Fact-checkers like Tardáguila said they will continue to push technology companies to do more, but at the end of the day, they acknowledged that these problems go far beyond Silicon Valley.

“False news is like water: You can't hold it; it goes through your fingers. It is just like lying. People will not stop lying. This is something we have to not only understand but accept. Lying is something that we do,” Tardáguila said.

Cover image: A woman holds a sign with an image of presidential candidate Jair Bolsonaro that reads "He lies in WhatsApp," during a protest against Bolsonaro in Sao Paulo, Brazil, October 20, 2018. REUTERS/Nacho Doce

This article originally appeared on VICE News US.