Bots spread fake COVID-19 info on Twitter ‘like it’s a propaganda machine,’ study says
How do you know if a tweet about the coronavirus is real? A new studyfound that almost half ofthe Twitter accounts posting messages about re-opening from the pandemic are software programs called bots.
Researchers at Carnegie Mellon University said their study found bots fueling conversations about reopening states, and some are spreading conspiracy theories and misinformation. The researchers found a subset of Twitter posts sharing conspiracies like linking the pandemic to 5G towers or hospital beds filled with mannequins.
“Conspiracy theories increase polarization in groups. It’s what many misinformation campaigns aim to do,” Carnegie Mellon Professor Kathleen Carley said. “People have real concerns about health and the economy, and people are preying on that to create divides.”
Spreading disinformation and conspiracies can make opinions more extreme and polarized, Carley said. “Increased polarization will have a variety of real-world consequences, and play out in things like voting behavior and hostility towards ethnic groups,” she said.
The university said Twitter bots are “dominating conversations about ending stay-at-home orders and ‘reopening America.’”
Researchers said they don’t know who is orchestrating the campaign.
“We do know that it looks like it’s a propaganda machine, and it definitely matches the Russian and Chinese playbooks, but it would take a tremendous amount of resources to substantiate that,” Carley said.
Twitter did not respond directly to the research, but in a blog post earlier this week, the company said it was actively working to identify and remove accounts that violate its policies.
The company said those policies include, “Malicious use of automation to undermine and disrupt the public conversation, like trying to get something to trend; Artificial amplification of conversations on Twitter, including through creating multiple or overlapping accounts; Generating, soliciting, or purchasing fake engagements.”
Earlier this month Twitter said it would start putting warnings and labels “that will provide additional context and information on some Tweets containing disputed or misleading information related to COVID-19.”
How to ID a fake Twitter bot
The university said researchers studied more than 200 million tweets about the pandemic. They found that 82% of the most influential retweeting accounts on Twitter are bots. “Of the top 1,000 retweeters, 62% are bots,” according to a university news release.
The researchers said they are continuing to collect tweets and are expanding their work to include posts on YouTube, Facebook and Reddit.
“We’re seeing up to two times as much bot activity as we’d predicted based on previous natural disasters, crises and elections,” Carley said.
Her team used artificial intelligence to figure out which accounts were run by bots, looking at factors like how often they tweeted, their followers and the network of mentions on the account.
“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,” Carley said.
“When we see a whole bunch of tweets at the same time or back to back, it’s like they’re timed,” Carley said. “We also look for use of the same exact hashtag, or messaging that appears to be copied and pasted from one bot to the next.”
The study found that 66% of the suspected bot tweets came from “accounts that are possibly humans with bot assistants.” Accounts that the researchers say are “definitely bots” sent out 34% of those Twitter posts.
Average Twitter users don’t need artificial intelligence software to identify suspect tweets. “Closely examining an account can offer indications of a bot, such as sharing links with subtle typos, many tweets coming out very quickly, or a user name and profile image that don’t seem to match up,” the university said.
“Even if someone appears to be from your community, if you don’t know them personally, take a closer look, and always go to authoritative or trusted sources for information,” Carley said. “Just be very vigilant.”
