A Carnegie Mellon University analysis found that 82 percent of the most influential COVID-19 retweeters are bots, and that bots comprise about half of the Twitter accounts that are crowing “Reopen America.”
We first got a sense there was something not-exactly-grassroots about the “Reopen America” movement back in April, when the New York Times reported that many of the state-level protests were funded by Education Secretary Betsy DeVos’ investment manager, Greg McNeilly. Soon-to-be failed Trump Federal Reserve nominee Stephen Moore flat-out admitted he was helping organize the Wisconsin protests, and arranging the payments for any arrestees' legal fees. He said in an April 14 Youtube video, “We have one big donor in Wisconsin, I’m not going to mention his name, he said ‘Steve, I promise I will pay the bail and legal fees for anyone who gets arrested,’” and then just to be more of a dick, added, “We need to be the Rosa Parks here, and protest against these government injustices.”
So clearly the “Reopen” movement is not exactly organic. But a new Carnegie Mellon study shows estimates that half the Twitter accounts discussing “Reopen America” are bots, that 82 percent of the most influential “Reopen” retweeters are bots, and that 62 percent of the top 1,000 retweeters are bots.
“We're seeing up to two times as much bot activity as we'd predicted based on previous natural disasters, crises and elections,” wrote computer science professor Kathleen Carley in the report. “Because it’s global, it’s being used by various countries and interest groups as an opportunity to meet political agendas.”
Carley notes that there are many different influence campaigns at work here, not all of these are MAGA Trumper accounts. Some are entirely committed to creating suspicion around 5G technology (perhaps to inhibit certain nations’ investment in that), others are anti-vaxxers enjoying their big moment with Bill Gates as the new Big Bad, and others are mere shysters hucking miracle cures.
"Tweeting more frequently than is humanly possible or appearing to be in one country and then another a few hours later is indicative of a bot," she added.
A Twitter spokesperson told NPR that they’d removed thousand of tweets, unveiled new “Potentially Harmful” tags for posts, and "challenged" 1.5 million accounts, whatever that means. But a spokesperson also challenged the definition of “bot,” saying the term could be used by "those in positions of political power to tarnish the views of people who may disagree with them or online public opinion that's not favorable."
The study’s authors say it’s an ongoing project, and that Facebook, Reddit and YouTube posts will be added into the analytical mix. But we can guess they’ll find the same misinformation tactics we saw in the 2016 election, as the “Reopen” movement looks to reopen some old wounds.
Related: Trump Is Already Trying to Lie About the COVID-19 Death Count; Kamala Harris Hopes to Stop Him [SFist]
Image: “Plandem¡c Part 1 [DELETED] | YOUTUBE CENSORSH¡P IS REAL” via YouTube