Automated Twitter accounts spread a significant amount of online misinformation during the 2016 election. A study in Nature Communications found bots were responsible for 34 percent of all shared articles from non-credible sources.
The spread of low-credibility content by social bots
By Chengcheng Shao, Giovanni Luca Ciampaglia, Onur Varol, Kaicheng Yang, Alessandro Flammini and Filippo Menczer
May 24, 2018
Researchers from Indiana University Bloomington found social media accounts that actively spread articles from low-credibility sources, or sources that “routinely publish various types of false and/or misleading news,” are significantly more likely to be bots. They analyzed 14 million messages that were spread from 400,000 articles on Twitter during and after the 2016 U.S. presidential campaign and election. Researchers identified two main reasons behind the bots’ success in spreading online misinformation:
- Twitter bots are especially active in the first few seconds after an article is posted on Twitter, which helps spread messages from low-credibility sources.
- Automated accounts specifically target influential users with a large audience through replies and mentions, which can manipulate users into sharing the content with their own followers.
Since low-credibility sources that become viral are heavily supported by automated accounts, this research suggests halting online bots could help to slow the spread of online information.
- From May of 2016 to March of 2017, researchers collected 389,569 articles from 120 low-credibility sites.
- Bots were responsible for 34 percent of all Twitter shares of articles from non-credible sources.
- Researchers collected data about the articles shared on Twitter through Hoaxy, an online platform that visualizes the spread of misinformation and fact-checking.
- In addition to these sites, researchers tracked 15,053 stories published by independent fact-checking organizations, including Snopes, PolitiFact, and FactCheck.org.