Mass Pro-Saudi Twitter Posts Are Suspicious
The Claim
The Diavolo network, ostensibly a network to automate repetitive tasks like copying and pasting, has made a series of claims across Twitter. Many tweets about Saudi Arabia contain similar language, saying that Saudi leadership has “always supported and called for moderation and fighting the [sic] terrorism.” Other tweets have included messages that Twitter ultimately decided to count as hate speech.
Claims were circulating across the MENA region.
News posted on
Emerging story
From 2015 on, Twitter has been inundated with accounts posting nearly-identical pro-Saudi Arabia messages. News sites including Al Jazeera wrote articles about the accounts.
Misbar’s Analysis
There are thousands of accounts across Twitter hosted by Diavalo that post similar content. Profiles associated with the accounts appear legitimate, with hundreds of tweets and pictures of real people. Many users retweet these messages, believing that they are being posted by real people.
However, in-depth investigations by Middle East Eye, Aljazeera, and The Guardian show that these tweets appear to be largely produced by bots.
The samples used by Middle East Eye show that of 20,000 tweets containing the phrase "Saudi 24 Analyst," nearly all of them were suspicious due to proximity of account creation dates, follower count, and content similarity. While Twitter has suspended many of these accounts, it has proven impossible to completely pause the bot infestation, primarily because it is so easy to sell fake followers. As one man contacted by Al Jazeera says, "I can create up to 2,000 followers per day, followers, retweets, website views. I sell all of it.” Internal analysis conducted by the Twitter safety team shows that this type of platform manipulaction is likely tied to state-backed forces. Since Diavolo is used largely for broadcasting (meaning its accounts do not interact), it is difficult to identify its bot army without relying on metadata.
Tweets about Saudi Arabia by accounts that do not interact with each other and use similar language are likely suspicious. Users should conduct their own ethical research and use critical thinking to identify if claims made by these tweets are true or false.
