According to research recently published by Indiana University and The University of Southern California, as many as 15 percent of active Twitter accounts are operated by bots. This means that many of the “followers,” “retweets” and “likes” that show up in your Twitter feed might not actually be from real people.
The research is likely to be raising a few eyebrows in Twitter headquarters, as the company is currently finding it difficult to expand its user base due to growing competition from Instagram, Minds, Snapchat, Facebook and more.
During the research at USC, over a thousand features were used as part of a framework to identify potential bots on Twitter. The features were selected from many categories including, duration between tweets, content of tweets, sentiment of tweets, friends, follower numbers and more. After analyzing all the data collected during the study, the researchers from the university estimate that 9% to 15% of Twitter accounts, which are active, are bots. Given the fact that Twitter has 319 million active users each month, the research suggests that as many as 48 million of those users are bot accounts.
In the report, researchers made it clear that the estimates were on the conservative side, since sophisticated bots utilizing the latest AI could have been falsely identified as humans under the framework model used in the study.
Twitter’s own estimate for active bot accounts is just 8.5%. However, my analysts are saying that Twitter would do well to take the 15% figure from this new independent report seriously, especially given the fact that the company’s revenue growth as dropped significantly. Twitter has really struggled to increase their active user base during the past 12 months.
Twitter has made is clear that it doesn’t think all bots are bad. In fact, a spokesperson from the company recently talked about how bot accounts can offer many benefits to users, as they can automatically issue warnings about natural disasters, for example.
Researchers from the universities responsible for the study also highlighted some of the benefits of bots, such as their ability to summarize news stories and alert people about the latest breaking news.
The USC report does recognize the many potential negative impacts of bots operating within the social media platform. For instance, it was noted that social bots could be used to promote terrorist propaganda, recruit extremists or drum up fake support for grassroots political movements.
Twitter provides users with the ability to report violations, such as spam posts and impersonation accounts. Some services also claim they can audit data and identify bot accounts.