The account, @WentDemtoRep, offered an online testimonial Sunday — the eve of a Republican convention featuring prominent African Americans challenging allegations of racism against President Trump — and was retweeted 22,000 times. Disinformation researcher Marc Owen Jones, of Hamad Bin Khalifa University in Qatar, found the tweet had 39,000 likes just 19 hours after it was posted.
Twitter removed the account Tuesday for policy violations after it had amassed 15,000 followers despite tweeting just a few times, all this month. The rapid spread of the tweet underscored how easily deceptive messages spread online — and how far they can get before social media companies are able to curb them.
“I’ve been a Democrat my whole life,” the tweet said, according to archived versions showing that the account was created this month, supposedly from Arizona. “I joined the BLM protests months ago when they began. They opened my eyes wide! I didn’t realize I became a Marxist. It happened w/o me even knowing it. I’m done with this trash. I’ll be registering Republican.”
Twitter suspended the account and several others that posted similar messages for violating rules about “platform manipulation and spam,” said a company spokesman, Trenton Kennedy. The company provided no other details of who created the account. In dozens of cases, the text from the original message was pasted directly into tweets of other accounts.
Jones, the researcher in Qatar, said, “These methods seem crude, but at the end of the day it shows how easy it is to game Twitter, and how a false account can get so many impressions and potentially influence or reaffirm the existing prejudices of an untold amount of people.”
Racial appeals and a focus on American racial unrest have been part of recent disinformation campaigns emanating from Iran, China, Russia and Romania, as detailed by independent researchers and major social media companies when they announced the removal of accounts that were fake or violated other policies against online manipulation.
Social media platforms have been too slow to respond to the use of their services to advance divisive or deceitful narratives about race, said Andre Banks, a co-founder of Win Black/Pa’lante, a group that combats disinformation targeting Black and Latino communities. He said technology giants appear not to “understand that cumulative impact or are afraid of challenging the stated intent of users where racial identity is a key part of distorting the truth.”
Russia’s Internet Research Agency, which U.S. officials indicted in 2018 for attempts to manipulate the 2016 election, borrowed heavily from the Soviet-era strategy of seeking to exploit American racial division to deepen unrest during the civil rights movement and after. Russian operatives portraying themselves as Black Americans — using account names such as “Blacktivist” and “BlackMattersUS” — were among the most prolific fake accounts used during that election and in the early months of the Trump administration, before they were uncovered.
“The IRA created an expansive cross-platform media mirage targeting the Black community, which shared and cross-promoted authentic Black media to create an immersive influence ecosystem,” a report for the Senate Intelligence Committee on Russian disinformation efforts said in 2018.
All but one of the 50 most widely retweeted posts by the Internet Research Agency were from fake accounts purporting to belong to Black Americans, according to Clemson University researchers Darren Linvill and Patrick Warren.
A common theme of such accounts were attacks on Hillary Clinton and other leaders of the Democratic Party, as well as the electoral process itself. Some posts directly urged Black voters to skip the election, such as a Facebook ad on Election Day targeting those with an interest in civil rights, the Rev. Martin Luther King Jr. and Malcolm X. The ad said of the presidential candidates, “Not one represents Black People. Don’t go to vote. Only this way we can change the way of things…”
An Iranian disinformation campaign, which was part a Facebook takedown in October, using a phony website, called BLMNews.com, and claiming to be a "Source of African-American News all around the world,” helped to push manipulative content, said Linvill, the Clemson researcher.
The most recent Facebook enforcement action, against behavior the company labeled coordinated and inauthentic, eliminated a network of accounts and pages originating in Romania and focusing on the November election, including by highlighting African American support for Trump. The misleading material even invoked the Obama family, suggesting the nation’s first Black president and his wife had endorsed Trump.
Last week, Facebook deleted a page using an image of LeBron James, among its deceptive tactics, to spread false and misleading claims about mail-in voting, a day after The Washington Post raised questions about the online operation. Ads purchased by the page were aimed at older users in Arizona, Nevada and Texas — among the competitive states where minority turnout could be decisive.
“Black voters are seen as a very necessary constituency for both sides. The scourge of racialized disinformation is really very strong right now because of that,” said Vanita Gupta, president of the Leadership Conference on Civil & Human Rights, a Washington-based umbrella group that has pushed the social media companies to police their platforms more aggressively for messages designed to suppress voter turnout.
The ambiguity surrounding the identity of the users was particularly pronounced in the case of the accounts that claimed to have forsworn the Black Lives Matter movement. Twitter revealed no details about the recently created personas beyond saying they had engaged in behavior contrary to the company’s policies.
That limited disclosure made it difficult to discern how the ideological work of the accounts was being carried out and by whom, said Deen Freelon, a professor at the University of North Carolina at Chapel Hill and the lead author of the recent article “Black Trolls Matter: Racial and Ideological Asymmetries in Social Media Disinformation.” He pointed to the “long history” of what the article terms “digital blackface” in finding that presenting oneself as a Black activist is the most effective disinformation tactic for driving online engagement.
At the same time, Freelon said, “authentic Black conservatives have long been trotted out by Republicans” in a bid to answer accusations of racism. If some of the users suspended this week by Twitter were expressing their genuine beliefs, he added, “the duplicate tweets posted by other accounts strongly suggest the kind of coordinated platform manipulation that Twitter expressly prohibits.”
Online platforms are particularly fertile ground for manipulating conversation among voters of color, opinion surveys suggest. Black and Hispanic users are more likely than White users to see social media as important to them in identifying like-minded people, according to the Pew Research Center. Pew’s poll found this summer that Black users are more than twice as likely as White users to say they have posted a hashtag associated with a political or social issue in the past month.