The internet, while an amazing avenue for like-minded individuals communicating about a diverse range of topics, is fostered primarily on anonymity, which is rather unfortunate in the murky concept of virtual space. There are dangers that be as one descends the rabbit hole that is the online web, and people often find it best to remain nameless, and even faceless when going online.
As it turns out, however, some of those anonymous folks might not even be real people at all. A new study has found that over 48 million Twitter users, about 15 percent of its total user base, are actually bots, meaning there's a good chance that different behaviors a user receives — such as favorites or retweets — may not even be the result of human action.
The research comes from the University of Southern California and Indiana University, and the findings of which could bring out a sore trouble spot for Twitter. It, at present, is already struggling to significantly expand the number of its users in light of the mounting competition from similar social media platforms such as Facebook, Snapchat, and a handful of others.
USC researchers took into account more than a hundred features to determine whether a Twitter account is actually a bot, according to CNBC. They used categories such as friends, tweet content and sentiment, and intervals between tweets. Leveraging this set of criteria, the searchers were able to estimate that between 9 to 15 percent of active Twitter users are bots.
Using USC's high-end estimate, that percentage amounts to nearly 48 million accounts out of Twitter's 319 million daily active users. However, the researchers state that some bots could have even been rendered as humans in the employed research model, skipping identification altogether, which according to the researchers makes the 15 percent figure a conservative estimate.
The research's estimates are greater than that of Twitter's own. The microblogging service in a SEC filing last month said that nearly 8.5 percent of all active accounts contacted Twitter's servers "without any discernable additional user-initiated action."
Not All Bots Are Useless
Not all Twitter bots, however, are useless, spam-spewing accounts. A spokesperson from the company said that many bots are actually beneficial, such as those which automatically caution people about natural disasters, or those which are related to customer service. USC's research even acknowledges this, citing that some bots are useful for disseminating crucial information such as news.
But while that's a relief, many bots are more malignant, with the research citing a ballooning record of Twitter bots that mimic human behavior to foster fraudulent grassroots political support and even ones which promote terrorist propaganda and recruitment.
Twitter currently offers a number of ways for users to report fake and spam accounts, but cracking down 48 million of them, excluding the beneficial ones, is quite a tall order, to say the least. On top of that, Twitter has a bigger problem of abuse, trolling, and hate speech proliferating its service, a noxious element of its site many critics have faulted it for not fixing.
Have you encountered any Twitter bots lately? Have they been useful or have they simply showed spam your way? Feel free to sound off in the comments section below!