Luisana Aguilar Alvarado contributed reporting.
It was early October 2019, and Quito was on fire.
Smoke from Molotov cocktails and tear gas blanketed buildings, mixing with the flames of burning tires. Whole families marched in the melee. Men with T-shirts tied around their mouths threw rocks at riot police, who responded with rubber bullets. There was a constant cacophony of sirens. This was the most violent protest in Ecuador’s recent memory. On October 1, President Lenín Moreno had announced that he would eradicate a 40-year-old fuel subsidy, cut some government workers’ salaries by 20%, and halve their annual vacation time, in exchange for a $4.2 billion loan from the International Monetary Fund to address an economic slowdown. Immediately, the price of diesel more than doubled. Many people were already struggling under growing inequality, and with this final blow, they took to the streets.
From the start, the president, flanked by soldiers, insisted that there were no real problems and accused his critics of orchestrating the protests. Simultaneously, his administration appeared to be stifling information. Spectators noticed that television channels looped mindless entertainment like “SpongeBob SquarePants.” Journalists from independent news outlets were scrambling to publish and transmit their work, but many found cell phone signals blocked. At least 16 reporters were attacked on the first day of the demonstrations, and the Inter-American Commission on Human Rights reprimanded the police for its brutality against the press. President Moreno’s interior minister, María Paula Romo, accused the media of vilifying the government in its coverage. At a press conference, she told a radio reporter, “You all have been very important for fake news.”
In the absence of real information, opinions formed fast on Twitter. But many weren’t taking aim at the government. Instead, social media overwhelmingly targeted the protesters.
There had been warnings that this would happen. In a 2015 report, the Psychosocial Research and Action Collective, an Ecuadoran research group, warned that “the government can convert into an ‘enemy’ anyone it wishes and distribute that massively.” Four years later, that prophesy appeared to be fulfilled. As the protests unfolded, the hashtag #CONAIEterroristas raced across Twitter, directed at the country’s most powerful activist group — the Confederation of Indigenous Nationalities of Ecuador, or CONAIE. “Now these savages have arrived with their lances,” wrote @CesarJavier_x19. “Absolutely every leader of CONAIE should be in prison for terrorism, kidnapping and damage to public and private goods,” wrote @nlopezcalle. Some people claimed that indigenous women were being used by indigenous men as human shields against the police; others said that Quito had been destroyed by indigenous activists who hated modernity.
The groundwork for this backlash had been put in place years earlier. CONAIE and other indigenous organizations were strong critics of former president Rafael Correa, who took office in 2007. Since the early years of his administration, fake social media accounts had consistently attacked them as backward, dumb, and violent. The racism indicated a fear of their power. There are 14 indigenous nationalities in Ecuador — together, about 1.1 million people — and they have a resounding political voice. Their activism around land and cultural rights has placed them at odds with the mining and oil-extraction aims of successive presidential administrations. It was only when CONAIE announced it would join the October protest that the crowd in the streets became a formidable force.
The unrest lasted 11 days. On October 13, when President Moreno finally negotiated a deal with indigenous groups to end their participation in the protests, the hashtag was repurposed to accuse the government of surrendering to terrorism. “This has got to be historic. A terrorist group blackmailing an entire country in real time,” wrote Twitter user @defreitasmario. “You already ceded to terrorism … don’t let them kidnap the entire country and impose their stupidity,” wrote @JF_Nicolalde, tagging President Moreno. And @MenendezOswaldo wrote, simply, “terrorism won.” On October 17, the hashtag #CONAIEterroristas (and close variants) reached its zenith, and it was tweeted and retweeted 8,310 times. In total, according to an analytics report, the hashtag was mentioned in 16,379 tweets and retweets that reached almost 5.1 million Twitter users.
To make sense of these numbers, social network researcher Alberto Escorcia ran an analysis of the hashtag using specialized software. The resulting image looked like the roots of a tree: Within the root system, there were clusters of squares, circles, and triangles representing bots that had coordinated mass retweets in short bursts of time. These bots didn’t interact with one another, and their identical behavior indicated that they were simply fine-tuned retweeting machines. “This is abnormal … because the way people debate and interact is through mentions or retweets that add something to the conversation,” said Escorcia. In addition to the bots, he also found clear evidence of troll accounts retweeting the hashtag.
Some users of the #CONAIEterroristas hashtag appeared to be real people who were perhaps unaware that many of their conversation partners were not humans but synchronized machines. The distinction can be subtle. Journalists and data scientists know to look for certain signs that a social media account is fake, and experts like Escorcia can use analytics tools to reveal evidence of irregular behavior. But there are no perfect ways to measure the impact of trolls.
The anti-CONAIE campaign was laden with an old racism. But the methods by which it was carried out were new, and were built on a strategy perfected by none other than former President Correa.
On January 24, 2015, Correa appeared on his weekly TV show “Enlace Ciudadano,” wearing a white button-up shirt and a smirk. A microphone hung jauntily from his left hand. The camera was positioned slightly below him, so viewers had to gaze up at him.
Correa did this show every Saturday morning during all ten years of his presidency, and because it was carried by at least 300 radio stations and two television channels, much of Ecuador heard or watched it. That day, he was talking about a citizen who had berated him on Twitter. The president’s staff had investigated the critic, and Correa was ready to reveal their findings to the nation. The offender turned out to be an 18-year-old boy. The president stated the boy’s full name and hometown and threatened him: “There are more of us [than you]. Many more of us.”
Then he addressed the nation. “For national security, we have to know who these sick people are,” he said. “Also to find out if, when they’re not anonymous anymore, they’re still brave enough to throw insults.” The crowd applauded and shouted “Bravo!” Correa asked citizens to report the real-life identities of any social media user who criticized him. He also encouraged them to counterattack.
This was aggressive political theater. But Correa didn’t need citizens to strike back. He had his trolls.
For most of his tenure, Correa’s approval ratings hovered above 60%. This was, in part, due to policies that seemed to stabilize the economy while easing inequality and a public relations strategy that made social media a battleground. Correa used state funds to set up an unknown number of fake Twitter accounts — at least hundreds, possibly thousands — controlled by what Ecuadorans called “troll centers.” He used three government ministries to engineer the misinformation, according to an investigation by the publication Códigio Vidrio published in April. This was one of the first instances of a politician mobilizing a Twitter army. They seeded scripts across the internet that promoted the president and attacked his enemies. “He knew how to get into your mind and heart,” says Efren Guerrero, a lawyer and political scientist who teaches at the Pontifical Catholic University of Ecuador. “The principal tool for Correa wasn’t public works, but instead communications.” And Twitter was key to this strategy.
The platform doesn’t reach the majority of Ecuadorans, but it is influential, a tool of those who run the country: urbanites and the political class. And Correa overrode the limits of Twitter’s natural audience in Ecuador by dragging his online battles onto his Saturday show.
In 2015, investigative journalists revealed the president’s use of trolls by publishing the details of contracts between the administration and a public relations firm. Those contracts were riddled with red flags and signs of backroom dealings. Public employees on ministerial communications teams began to whisper about being forced to tweet in support of the administration. Trolling had appeared in Ecuador before Correa was elected, said Apawki Castro, the communications director of CONAIE, but it wasn’t integrated into politics — and it didn’t look anything like what it later became.
Although trolls maintain exaggerated and aggressive public personae, the people behind these accounts tend to disappear when approached directly. No former or current government employees forced to engage in abusive online behavior agreed to go on the record for this article. But someone has to deal with the wreckage they leave behind. As in much of the world, there are now digital privacy organizations across South America that do everything from facilitating training sessions on digital rights to documenting attacks to supporting victims.
All told, we will never know how many troll centers attacked Ecuadorans for Correa or how many employees worked in them. But it was clear that Correa was their leader.
After that Saturday show in 2015, a Twitter army went after the 18-year-old critic. It was a relatively demure display of their capability; attacks could be far larger. They often happened during the Saturday morning show, when troll accounts would turn Correa’s thoughts into trending topics. “It was a tweet war,” says Roberto Chávez, a journalist at the progressive Ecuadoran news outlet Wambra, who would watch these play out in real time. But beyond the Saturday swarms, the trolls had other duties. “Correa would send out five tweets on Mondays at 8 a.m., and that was the political agenda he was going to talk about that week,” said Chávez. “There were likes and retweets, immediately. The messages were repeated: ‘Yes, my president. Thank you for everything, my president.’”
When the task was a hatchet job, the targets were often journalists, indigenous people, and feminists. One of the countless times that Carlos Andrés Vera, a filmmaker, was attacked, aggressors all tweeted the same photograph at him. It was an image Vera had taken of his children on his phone that he had never published anywhere. “The message was ‘We have access to your phone. We’re watching you,’” says Vera. “Obviously, you’re not used to that, and it gives you anxiety.”
From there, the abuse only escalated. Vera, who has made documentaries about indigenous people and was a vocal critic of Correa’s mining and oil drilling in indigenous territories, was denounced as an enemy of the state. Once, a troll threatened to make a pornographic video of his young son. Vera now has 9,080 accounts blocked on Twitter. He used to imagine trolls as “psychopaths snorting cocaine, doubled over their keyboards,” but now he knows better. “The government became that psychopath,” he says. “That’s the big leap that Ecuador made.”
Anyone who publicly criticized the president risked being attacked. The exact location of a feminist leader’s home was published on Twitter. Investigative journalist Martha Roldós had her email account hacked, and Correa read from its contents on his show. When lawyer Silvia Buendía criticized the president’s opposition to LGBT and reproductive rights, trolls sent her death threats.
That such harassment would win public support might have been hard to understand before the most recent presidential elections in countries like the United States, the Philippines, El Salvador, and Brazil. Correa, like other populist leaders, positioned himself as an outsider bucking the system. Some Ecuadorans expressed rage at his online tactics. But his redistributive policies improved many lives, and in that light, online aggression may have seemed like a footnote in a historic battle against inequality.
And, as in other countries, once trolling was injected into public discourse, it became normalized.
When Correa’s term ended in 2017, he left the presidency, and the Ecuadoran internet went relatively quiet. A number of his former victims believed he had taken his trolls with him. But this didn’t last long. The practice of trolling remained, and it spread. Hiring trolls, and acting like one, became part of wielding presidential power in Ecuador. Last summer, Twitter announced it had dismantled a network of 1,019 mostly false accounts linked to the current administration. Moreno appears to have far fewer trolls than Correa, though employees of ministerial communications departments still complain in private WhatsApp groups about being expected to tweet in support of the administration. Of course, Moreno’s government is operating under austerity measures. It’s possible he simply can’t afford an army the size of his predecessor’s.
Trolling was a major success for Correa, and it has since become a strategy for mobilizing around any political cause. “After Correa, the practice of having trolls broadened and democratized,” says Martha Roldós, an investigative journalist who also founded Fundación Mil Hojas, the publication that uncovered the administration’s trolling contracts. Correa and his acolytes still sometimes mobilize the army; Código Vidrio’s investigation unearthed its presence during the October 2019 protests and in response to the Covid-19 pandemic. But troll victims noted that since Correa’s departure, there has been a proliferation of fake accounts that seemingly represent interest groups — anti-abortion organizations, for instance — with no apparent connection to the government.
One independent troll attack came in the form of a low-budget video that appeared on Twitter during a 2019 legislative battle over abortion rights. The video splices photos of pro-choice and LGBT leaders into a shadowy scene of a witch brewing her potion. Feminist lawyer Soledad Angus Freré, who appears in the video, says she’s now used to such attacks, which she is convinced originate in “anti-rights troll centers” run by private right-wing groups. Many victims of post-Correa trolling echo this belief. For instance, when, last year, Wambra published an investigation into the plight of girls younger than 14 who had been forced to give birth after being raped, journalists noted a trend in Twitter reactions: a known conservative social leader would attack tweets promoting the investigation, and a wave of fake accounts would immediately echo the attack.
Buendía, the lawyer who received death threats after criticizing the president on Twitter, experienced this culture shift personally. “When Correa left the government, I saw the attacks absolutely lower in intensity,” she says. “Now there are different types of attacks, much less formal, less professional,” orchestrated by “real people who have fake accounts, and they use them to accost and humiliate people,” she said. “It’s not like in Correa’s time — they are far fewer and less organized, but they’re equally or more violent.”
In one way, they’ve become more sophisticated. Troll accounts used to be easy to identify. They had been recently created and usually had few followers, who often followed each other. Their handles were typically random names followed by meaningless numbers. If they had any profile image at all, it was often something like an inspirational quote or a flower. “It was so clumsy,” says Valeria Betancourt, who analyzes information policy at the Association for Progressive Communications in Quito. “It was so clear they were trolls.”
But as these accounts aged, they began to seem more authentic. They accrued long tweet histories spanning various topics. The people behind them grew more clever. They stole real photos to use for their profiles and came up with better handles. Practices like “account trafficking,” in which real people sell their accounts for trolling use, made it worse. Now, “at first glance, they appear to be real,” says Vera.
Perhaps what is most striking is that the strategy came full circle: While troll accounts strove to appear credible, ordinary citizens began adopting troll-like behavior, mimicking the aggressive and violent language of bots. Alexis Moncayo, a radio journalist who covered the protests last October, was bombarded by a swarm of trolls a week after the protests stopped. Attackers, whom he believed were aligned with the Moreno administration, accused him of “inciting rebellion” and conspiring with former president Correa. Then the in-person intimidation began. Two residences on the ground floor of his apartment building were robbed in broad daylight. A political figure sent him threatening messages on WhatsApp. He started regularly inspecting his car in case drugs had been planted in it. Ultimately, at the insistence of his family, he left his job.
The former administration may have managed its trolls, but no one manages trolling itself. Once introduced into an environment, trolls will fundamentally alter it, and it’s not clear if that corrosive process can be reversed. In hiring people to act as trolls and paying them with government money, Correa changed how debates about beliefs, policies, and candidates play out in Ecuador.
Last October, the lawyer Lenin Sarzosa found himself in court defending 33 indigenous people who had been arrested in that month’s protests. Six were minors, and all were facing charges of terrorism. The prosecution alleged that they had set a government building on fire during the turmoil, but before long, the terrorism charges were dropped — there was no evidence to back them up. Even so, the case was a communications victory for the trolls. That hit Sarzosa when he saw the hashtag #CONAIEterroristas graffitied on a wall in Quito. The slogan spread fast — online trolling had merged with the real world.
Guerrero, the political scientist, believes that online abuse will persist as a political strategy. “The medium is the message. Even the president has become a public relations product,” he says. “So you grab the internet, and you manipulate it by poisoning information.” He predicts that candidates in the country’s 2021 presidential election will need their own stables of trolls in order to be competitive.
Because presidential trolling exploded in Ecuador before it did in other countries, Guerrero describes the country as a bellwether for developments in online bullying and fake news: “Ecuador has been a test tube. It’s a tiny country that has been a big, big experiment.” National media have always been a force for presidents to reckon with, Guerrero says, but Correa simply bypassed them and used social media to devise the headlines he wanted. Even when he was criticized for it, he still controlled the conversation. Guerrero points to Correa’s continuing popularity and his 3.5 million Twitter followers. “We had fake news before the concept of fake news existed,” he says. “People in the U.S. are surprised by Trump, but for us — been there, done that.”