MIAMI — False claims on WhatsApp and Telegram that monkeypox isn’t a real virus. Russian war propaganda reaching from Latin America to Hispanic voters in the U.S. Unfounded assertions that the FBI’s raid of Mar-a-Lago was ordered up directly by President Joe Biden.
Two years after controversy erupted over the influence Spanish-language disinformation had on U.S. elections, the problem persists on social media channels and in mass media.
Tech companies and the federal government say they are trying to address the issue as yet another election season gets closer. But they acknowledge that they’re wading through a flood of information even as they are urged to do more to regulate inaccurate and manipulated Spanish-language content — which some lawmakers and advocates say continues to be under-regulated on social and mass media.
“Although Facebook and the other platforms have been under a fair amount of pressure to eliminate and contextualize disinformation on their platforms, their focus has been almost entirely on English language material,” said Tony Affigne, a Latino politics professor at Providence College. “So they’ve allowed a great deal of Spanish language disinformation to slip through the cracks.”
To better understand the latest efforts to regulate Spanish-language content, the Miami Herald and el Nuevo Herald sought answers from the federal government and the world’s largest tech companies.
One thing seems clear: They say they’re making specific efforts to counter disinformation affecting Spanish-speaking communities.
— The Federal Communications Commission, which works on a complaint-driven basis to review “media distortion” on radio, TV, satellite and cable in all 50 states, has never investigated a case in Spanish, a spokesperson told the Herald. But a different federal agency, the Federal Trade Commission, is taking steps to regulate online content that some advocates believe could help curb Spanish-language disinformation.
— Twitter started working in August with the Spanish Language Disinformation Coalition ahead of the midterms to help the platform take action against misleading claims that dissuade people from participating in the election process.
— TikTok’s in-app election centers are available in dozens of languages, including Spanish, to provide access to information about elections and voting by BallotReady and the National Association of Secretaries of State. TikTok is, according to the company, “extremely confident” in its moderators’ ability to moderate Spanish-language content.
— Meta, Facebook’s parent company, has partnered with five fact-checking organizations to identify, review, and rate content in Spanish. Every time a fact-checker rates a piece of content as false, Meta reduces its distribution so fewer people see it. The company, which owns WhatsApp as well, has also partnered with Telemundo and Univision to launch fact-checking tiplines in Spanish on the platform, according to a Meta spokesperson.
“In addition to Spanish-speaking fact-checking partners, we have native Spanish-speaking content reviewers who are based in the U.S.,” a spokesperson for Meta said, adding that those efforts are assisted by artificial intelligence that is capable of learning slang, colloquialisms and other idioms. “Spanish is one of the most common languages used on our platforms and is also one of the highest-resourced languages when it comes to content review.”
A flood of content
Since the beginning of the pandemic, Meta has removed more than 24 million pieces of content from Facebook and Instagram globally. The company has also removed over 3,000 accounts, pages, and groups for “repeatedly” violating their rules against spreading dangerous COVID-19 and vaccine misinformation.
On YouTube, about 1 of every 1,000 views involves content that violates the company’s community guidelines, a spokesperson said, though a recent internal analysis found that 75% of problematic videos were removed from the platform before getting more than 10 views.
YouTube says it has more than 20,000 people around the world, including many with Spanish-language expertise, who work to detect, review and remove content that violates their policies. Google News Initiative, according to a spokesperson, is working with a fact-checking group based in Latin America called LatamChequea to “train 500 new fact-checkers in Argentina, Colombia, Mexico and Peru.”
The reach of Spanish-language media
Because so many U.S.-based Latinos rely on social media as a way to stay connected to their family and friends across Latin America, it can make them more susceptible to misinformation online. YouTube and Meta remain the most popular with Hispanic users online, with WhatsApp being used more by Latinos than any other demographic in the country.
Hispanic YouTube users spend twice as much time on the platform compared to non-Latinos, according to research by Equis. YouTube is also used by 85% of adults who identify as Hispanic, a larger percentage than any other demographic, according to recent Pew Research.
“We can see in our own families and communities that people are being affected,” said Affigne, the Providence College professor, who has argued that users in Florida and Texas are among the most targeted. “People are saying that they saw on the internet that the COVID-19 vaccine has microchips, that President Biden has been diagnosed with dementia or that monkeypox is an airborne virus. These are dangerous ideas that are costing people their lives.”
Darren Soto, a Central Florida Democrat and member of the Hispanic Congressional Caucus, said increasing Spanish moderation and the amount of people who monitor this content is going to be key for the 2022 election.
“We’re not talking about debates between progressive and conservative voices, we’re talking about dangerous lies like who won the 2020 election or the health benefits of the vaccine,” Soto said. “All these go beyond the normal traditional debates and many of them are unlawful.”
Pressure to do more (and less)
Still, efforts to tackle misinformation are not without controversy. Conservatives argue that content labeled disinformation is often simply political opinion that is unpopular on the left, or in some cases found later to be true when additional facts emerge.
The American Civil Liberties Union has also warned of overreach by the government and social media networks: “We’re skeptical of the government arbitrating truth and falsity,” the organization tweeted after the Department of Homeland Security announced the creation of a Disinformation Governance Board that was quickly disbanded amid criticisms.
It’s a tricky position for both the social media companies and the government, which are simultaneously being asked to do less and more.
Hispanic members of Congress, for instance, argued in May that Spanish-language disinformation is “running rampant” on encrypted messaging apps like Telegram. The following month, U.S. Sens. Bob Menendez, Bill Cassidy and Tim Kaine called on Meta, Twitter and Telegram to address Spanish-language war propaganda by Russian media outlets that they said was spreading from Latin America to users in the U.S.
For example, RT en Español, the Spanish-language channel of Russia’s global TV network based in Moscow, posted a video statement on its Facebook page, which has 18 million followers, challenging the facts of the June attack on a Central Ukraine shopping mall amid Russian President Vladimir Putin’s invasion of the neighboring country.
“We are deeply concerned by reports that the operations and reach of such outlets have only increased amid Putin’s actions in Ukraine,” they wrote.
Menendez and U.S. Rep. Tony Cárdenas, D-California, are exploring what their next steps are going to be given the response they’ve received from Meta and the lack of response from Telegram and Signal, a spokesperson for Menendez said. The Congressional Hispanic Caucus is seeking a meeting with the heads of these social media companies to see what their efforts to curb Spanish disinformation will look like.
In many ways, platforms are playing a game of whack-a-mole.
Following the FBI’s search of Mar-a-Lago, conservative radio personalities and politicians began spreading narratives on social media and radio comparing the FBI to “secret police agencies in Latin American authoritarian regimes”— an argument legal experts and former prosecutors say ignores the strict legal process behind the application and authorization of the warrant.
On Twitter, Americano Media, a new Spanish-language conservative outlet on satellite radio, referenced comments to Fox News by a station executive, Michael Caputo, claiming “There is no difference between the FBI and the KGB. We are living in Russia!!!!”
La Derecha Diario, a right-wing Argentinian publication with over 173,000 followers on Twitter, stated: “Biden freaked out. He sent the FBI to raid Trump at Mar-A-Lago, broke into his safes, and took his computers from him.”
More recently, right-wing and conspiracy theory-focused YouTube channels have started to translate content from English-language right-wing media in an effort to spread the same information to Spanish speakers, “thus exposing them to narratives they may have otherwise not seen,” according to Media Matters, a left-leaning non-profit dedicated to fighting misinformation.
Skepticism
Evelyn Pérez-Verdía, a Florida-based Democratic consultant and the chief strategy officer of We Are Más, a firm focusing on countering disinformation, said social media networks need to be more transparent about the resources they’re putting into fighting disinformation. She remains skeptical of their efforts.
“They should be working with the people already working on fighting Spanish disinformation. And if they are, we’re not aware of it,” she said of tech companies.
In testimony before Congress last October, Facebook whistleblower Frances Haugen revealed that 87% of misinformation spending on the platform is on English-language content, but only about 9% of the users are English speakers.
And during the height of COVID-19, the online activist group Avaaz found that significant delays in Facebook’s implementation of its anti-misinformation policies resulted in millions of users seeing “harmful misinformation content,” and that the company also failed to issue warning labels on 70% of misinformation in Spanish, compared to 29% in English.
The government also faces continued criticism.
Due to First Amendment sensitivities, and because claims of disinformation in broadcast media must show that there is “deliberate intent to distort or mislead an audience,” the FCC rarely takes action over manipulated content, according to a spokesperson. The last case investigated was in 2007, and involved an English-language station.
“They’re getting away with murder on air,” Raul Martinez, the Democratic former mayor of Hialeah and host of a YouTube talk show, said about the hosts of conservative Spanish-language radio shows in Miami.
And while the Spanish Language Disinformation Coalition wrote a letter in June to the FTC to thank the agency for taking additional steps to curb manipulation of social media platforms, its directors also said that the problem is far from contained. The FTC is considering creating a rule that would “curb lax security practices, limit privacy abuses, and ensure that algorithmic decision-making does not result in unlawful discrimination.”
This type of act is necessary because internet platforms thrive based on business models which extract data to collect, distribute, and leverage users’ personal information across their services, the Spanish Language Disinformation Coalition said in a letter to the FTC.
“While we can see some progress from platforms on English-language content, it’s common for Spanish-language posts that clearly violate the platform’s community standards to stay up for months,” they wrote.
———
(Lesley Cosme Torres stories for the Miami Herald are funded by a grant from the Knight Foundation. The Miami Herald retains all editorial control of the content.)
———
