Botnets have been an ongoing headache for Twitter, and now a botnet with over 350,000 bots — nearly rivaling the Srizbi botnet — has been discovered, and it really loves Star Wars. (More on that in a moment.) Bots often have no followers, few tweets, and use the default photo of a cartoon egg as their profile picture. But as pointless as that sounds, they can be pernicious little bastards, engaging in spreading fake news as well as spamming, astroturfing, manipulating trending topics and perceived public opinion, contaminating streaming API (which damages research work), and selling fake followers.
In worst case scenarios, bots could even spread propaganda that results in someone shooting up a pizza place or in politicians threatening nuclear war. That seems kinda bad.
University College London researchers Juan Echeverria and Shi Zhou discovered a botnet with unique features that might make it hard to detect by current cybersecurity methods. They were able to find the bots because their geo-tagged tweets had an odd distribution (sometimes tweeting from uninhabited places) and used the same set of books as their English language reference. Yes, this botnet loves Star Wars.
To appear fluent in English, every bot tweeted fragments of text culled from eleven different Star Wars novels, occasionally adding a hashtag (like #teamfollowback). Do they just really like reading about Leia using the force and Han saving Chewie? And is this why an egg told me I was thin as a slip of paper and just as useless?
It’s a mystery. The botnet has been dormant since its creation in 2013 and exists for unclear reasons. Hopefully, these bots will never be used for, say, trying to start a war or encourage a vigilante shooting. If this Star Wars-loving, 350,000 strong army of bots is unleashed someday, we can only hope that their sole purpose is merely to swear at us like the other Twitter eggs do.
https://twitter.com/jhonrules/status/637130917664813056
(Via Gizmodo, Discover Magazine, and ArXiv)