(WASHINGTON) — Coordinated groups of inauthentic accounts have been attempting to influence online conversations around the 2024 elections for the better part of a year, according to a recent investigation conducted by Cyabra, a social analysis firm.
According to the Israel-based firm, someone created thousands of automated Twitter accounts that appear to be praising former President Donald Trump and criticizing his political rivals on both sides of the aisle.
“What I found was so interesting about this bot farm [was] that it understood the nuances of the division within the Republican Party and it was exploiting that online,” Jules Gross, a solutions engineer at Cyabra, told ABC News.
Gross said she reviewed hundreds of posts and found that these automated accounts — or bots — were going after Trump’s potential 2024 rivals for the presidency.
As soon as Nikki Haley announced a bid for the presidency in February 2023, Gross said she saw bots accusing Haley of being disloyal to her onetime boss.
Around 40% of the conversation online about Haley was being controlled by inauthentic accounts versus just 2% before her announcement, according to Cyabra.
Haley hasn’t been the only target; these bots also appeared to go after Florida Gov. Ron DeSantis, according to the firm.
Although DeSantis hasn’t announced a bid, it hasn’t stopped numerous fake accounts from insisting he not run in the 2024 election, Cyabra said in its report.
Gross said she has also found hundreds of posts accusing Republicans of not being loyal because they cooperated with Democrats.
“Whether it’s Kevin McCarthy, or even Paul Ryan, Mitch McConnell and also Nikki Haley, the bots say these people are not Republican enough,” Gross said.
She added that the bots allege that “(t)hey are not true Republicans, they need to be more extreme.”
Gross said Cyabra found three coordinated groups of inauthentic accounts created on one of three dates in April, October and November of 2022.
Gross said she was initially scanning conversations around prominent American politicians on Twitter when she noticed seemingly inauthentic accounts popping up among real ones.
“Once I filtered all of the fake profiles from the ones that were just created, on the same days, I could see that they were all promoting the same ideology online, possibly even posting the same exact posts online,” Gross said. “And from that, we’re able to deduce that is a coordinated activity.”
While public perception may be that social media bots are rudimentary, some experts say these coordinated groups are getting more complex and harder to spot.
“Detecting bots is a very difficult challenge that requires analysis of a number of different parameters,” said Sam Woolley, the program director at Propaganda Research Lab at the University of Texas, Austin. “You can’t just look at two or three different things.”
Cyabra said it looked at over 500 different behavioral parameters to identify whether these profiles were fake — parameters like how many hours a day an account was active, the geographic location of their followers and if the content posted was original or repurposed.
As of Thursday, the three coordinated groups identified are still active on Twitter, Gross said. Most recently, Cyabra said these accounts have been posting and re-sharing posts about the health and whereabouts of Democratic Sen. John Fetterman of Pennsylvania following his recent hospitalization for severe depression.
According to Cyabra, Elon Musk hired the data analysis firm to analyze bot activity on Twitter before he purchased the company.
ABC News has reached out to Twitter for comment.
“It shows just how likely it is that the conversations on social media surrounding the 2024 presidential election will be controlled and will be full of not just of disinformation and misinformation, but also harassment, disenfranchisement content, and all other sorts of forms of misleading information,” warned Woolley.
Cyabra said it does not know who is behind these coordinated accounts.
Research found that bots have helped spread falsehoods and manipulate conversations in all major U.S. elections — and many others worldwide — since 2016. Bots spreading misinformation aren’t specific to Twitter; networks of inauthentic accounts have been uncovered and taken down on Facebook and Instagram in the past, experts say.
But experts like Woolley say social media companies need to “take a stand and provide more clarity with what kinds of automation is unacceptable.”
Copyright © 2023, ABC Audio. All rights reserved.