When the gunman who attacked students and staff at Marjory Stoneman Douglas High School in Parkland, Florida, this February stopped shooting, 17 people were left dead or fatally injured. The outcome was devastating.
To add insult to injury, shortly after the tragedy, Twitter bots began sending out emotionally charged quips and conspiracy theories about what took place. However, a retrospective analysis of those tweets suggests that humans were the ones who spread them furthest.
A team at the University of South Carolina downloaded seven million tweets, posted during a one-month period shortly after the shooting, and isolated tweets originally posted by 400 bot accounts. Only automated accounts that appeared to be trying to influence people鈥檚 opinions were included, rather than, for example, benign bot accounts used by news organisations to publish links to stories.
Advertisement
The malicious bots sometimes sourced their content from untrustworthy or politically extreme websites or appeared to simply make up provocative statements. But it was the human response to this content that really got the snowball rolling. Over 90 per cent of retweets given to bots were from human-run accounts.
About one-third of the tweets shared in this way could be described as attempts to 鈥渂ait鈥� users by targeting their beliefs, says , who worked on the project. For example, one popular tweet referred to 鈥済un control dolts鈥� while another described CNN as 鈥淢arxist鈥�. It is not clear who was behind the bots tweeting about Parkland.
Such content may act as a 鈥済ateway drug鈥�, pulling Twitter users towards polarised news sources or ideologies, says Kitzie.
Automated accounts succeed when they trick both humans and social media website algorithms, which select popular content, into sharing their messages far and wide, says at the University of Indiana.
鈥淎ny kind of manipulation is bad, but certainly manipulation that exploits this kind of event is even more鈥� disgusting, atrocious,鈥� he says. 鈥淭hat鈥檚 why we study this stuff 鈥� because we want to make it more difficult for people to manipulate social media.鈥�
Reference:
