A new study by researchers at the University of Southern California found that bots are influencing the US election discussions on Twitter. On more than 240 million election-related tweets, the co-authors identified thousands of automated accounts that had tweets about President Donald Trump, his Democratic opponent, former Vice President Joe Biden, and both campaigns. Most of the bots promoted right-wing political conspiracies like QAnon and politically biased narratives about the origins and treatments of the novel coronavirus. And while it is believed that the bots were responsible for a few million of the tweets, the researchers found that they potentially reached hundreds of thousands of users on Twitter.

The results seem to confirm the fears of some social media pundits who have voiced concerns that bots themselves will dodge elaborate filters to amplify misleading information, disrupt voting efforts, and create confusion after the elections. The role that bots play in spreading false and misleading information is well known. Research by Indiana University scientists found that over a 10-month period between 2016 and 2017, bots addressed influential users through responses and mentions to reveal untrue stories before going viral. During the 2017 Catalan independence referendum in Spain, bots generated and promoted violent content aimed at users demanding independence.

The University of Southern California study, published Wednesday by the online journal First Monday, focused on election-related tweets from June 20 to September 9, 2020 and other data from Twitter during that period. During their analysis, the co-authors identified significant differences between bots and humans, as well as the type of voting content they tweet and retweet on the platform. In addition, they examined the political leanings of real human users, popular hashtags and tweets that included stories or other content from partisans and traditional news media.

Bots almost exclusively retweeted original posts on Twitter by human users, according to the report’s co-authors. In return, many people retweeted the bots’ messages that matched their political leanings. Right-wing accounts outperformed their left-wing counterparts by 4 to 1 for bots and 2 to 1 for humans. Meanwhile, users who shared or retweeted messages from right-wing media platforms were almost twelve times more likely to be conspiracies than users promoting left-wing content (25%) compared to 2%). Only 4% of bots retweeted messages from left-wing and centrist media like the Washington Post, New York Times, Los Angeles Times, ABC News, BBC, CNN, and others, while about 20% of users shared content from right-wing media like Breitbart, OANN and Infowars were probably bots.

One of the conspiracy theories the researchers tweeted was “pizzagate,” a debunked claim linking Democratic Party officials and US restaurants to child trafficking. Approximately 13% of all users exchanging conspiracy reports were suspected bots, including several from Ghana and Nigeria, who launched information campaigns to educate left-wing users about the Black Lives Matter movement. Saudi Arabia and Turkey also had high levels of engagement with right-wing users, while Russia and China mainly targeted left-wing fringe and conservative groups.

Twitter previously denied the notion that third-party services could accurately detect bot activity without access to its internal records. However, the findings by Ferrara and colleagues are in line with other research into bot activity on the platform so far. A Carnegie Mellon team found that bots can account for up to 60% of the accounts discussing COVID-19 on Twitter, advocating false medical advice, conspiracy theories about the virus, and efforts to end bans. And Bot Sentinel, who is tracking bot activity on social networks, watched new reports of disinformation campaigns on Black Lives Matter in early July, including false claims that billionaire George Soros was funding the protests and that George Floyd’s murder was a hoax.

The Audio Problem: Learn How New Cloud-Based API Solutions Solve Incomplete, Frustrating Audio In Video Conferencing. Access here