Fake US Election Accounts Proliferate on X

A social media analysis company’s report, obtained with Reuters confidentially before its Friday release, found that fake accounts discussing the U.S. presidential election are flourishing on X

A subset of artificial intelligence known as machine learning is utilized by analysts at the Israeli technology company Cyabra to identify phony accounts.

The company discovered that 15% of X accounts that laud former President Donald Trump and criticize President Joe Biden are fraudulent. Additionally, the report revealed that 7% of accounts that extol Biden, a Democrat, while disparaging Trump, a Republican, are fraudulent.

The study by Cyabra is predicated on a two-month analysis of tweets from the X platform, formerly called Twitter, commencing on March 1. The evaluation encompassed examining trending hashtags and classifying posts as positive, negative, or neutral in sentiment. 

According to the analysis, newly identified fraudulent accounts experienced a tenfold surge in March and April. 

The report identifies 803 inauthentic pro-Biden profiles out of 10,065 and 12,391 inauthentic pro-Trump profiles out of 94,363.

No spokesperson for X nor representatives from the White House or the Trump campaign responded to a request for comment regarding the phony accounts. 

Social media platforms, including X, have come under increased scrutiny ever since Russia interfered in the 2016 U.S. presidential election to bolster Trump’s candidacy and undermine Democratic Hillary Clinton.

Election officials and experts in online misinformation are once more vigilant in search of deceptive narratives in anticipation of the November 5 election.

Cyabra stated that the fabricated accounts endorsing Trump during this cycle are components of a coordinated effort to manipulate online discourse and sway public opinion. The report did not disclose the identities of the individuals or organizations responsible for the campaign. 

Cabrera stated that it arrived at this conclusion by analyzing evidence such as the concurrent publication of posts and remarks by fake accounts and the use of identical hashtags. The phony pro-Trump accounts, according to the report, propagated two primary arguments: “Vote for Trump” and “Biden is the worst president in U.S. history.”

“The degree of coordination implies the presence of a malicious intent and a comprehensive scheme to sway public sentiment,” stated Rafi Mendelsohn, vice president of Cyabra. 

The report stated that the phony accounts endorsing Biden were separate from a coordinated campaign because the characteristics of such an operation, such as simultaneous fake account posts, could not be identified. 

X, a publicly traded company until its acquisition by billionaire Elon Musk in 2022, has consistently denied the existence of fraudulent accounts on its platform.

In May 2022, Twitter reported that an internal evaluation of accounts revealed that less than 5% of its daily active users were “false or spam.” Cyabra estimated at the time that 13.7% of Twitter profiles were fraudulent. 

Elon Musk announced on April 4 via X that a “system purge of bots and trolls” was in progress and that the organization “will be employing the full force of the law to apprehend the individuals accountable.” To combat spammers and bots, the company piloted its “Not a Bot” program in New Zealand and the Philippines in October.