New study by USC researchers shows bots evolving to better mimic humans during elections.
USC Information Sciences Institute (USC ISI) computer scientist, Emilio Ferrara, has new research indicating that bots or fake accounts enabled by artificial intelligence on social media have evolved and are now better able to copy human behaviors in order to avoid detection.
In the journal First Monday, research by Ferrara and colleagues Luca Luceri (Scuola Universitaria Professionale della Svizzera Italiana), Ashok Deb (USC ISI), Silvia Giordano (Scuola Universitaria Professionale della Svizzera Italiana), examine bot behavior during the US 2018 elections compared to bot behavior during the US 2016 elections.
Over 30,000 bots were found by the researchers after they examined almost 250,000 active social media users who discussed the US elections in both 2016 and 2018. They found that bots in 2016 were primarily focused on retweets and high volumes of tweets around the same message. However, as human social activity online has evolved, so have bots. In the 2018 election season, just as humans were less likely to retweet as much as they did in 2016, bots were less likely to share the same messages in high volume.
Bots, the researchers discovered, were more likely to employ a multi-bot approach as if to mimic authentic human engagement around an idea. Also, during the 2018 elections, as humans were much more likely to try to engage through replies, bots attempted to establish voice, contribute to dialogue, and engage through the use of polls, a strategy typical of reputable news agencies and pollsters, possibly in an effort to lend legitimacy to these accounts.
In one case, a bot account asked on Twitter if federal elections should require voters to present ID at the polls. It then requested that Twitter users vote and retweet.
Lead author, Emilio Ferrara, noted, “Our study further corroborates this idea that there is an arms race between bots and detection algorithms. As social media companies put more effort to mitigate abuse and stifle automated accounts, bots evolve to mimic human strategies. Advancements in AI enable bots to produce more human-like content. We need to devote more effort to understand how bots evolve and how more sophisticated ones can be detected. With the upcoming 2020 US elections, the integrity of social media discourse is of paramount importance to allow a democratic process free of external influences.”
Reference: “Evolution of bot and human behavior during elections” by Luca Luceri, Ashok Deb, Silvia Giordano and Emilio Ferrara, 2 September 2019, First Monday.