The Dutch far-right organized an inflammatory cartoon contest involving Islamophobic material and bots joined the backlash
On June 12, Dutch far-right politician Geert Wilders announced that his Freedom Party will hold a competition for cartoons depicting the Prophet Mohammad. It immediately caused an outcry among Muslim communities as any visual depiction of the prophet is considered to be blasphemous in the Islamic tradition.
Wilders and his party have openly derided Islam and its followers for years. In 2017, before the Dutch parliamentary elections, he referred to Moroccan migrants as “scum” and said, “Islam and freedom are incompatible.” Last March, Wilders was accused of spreading hate speech and incitement to violence against Dutch Muslims after his party aired a campaign video referring to Islam as “terror,” “slavery,” and “deadly”.
Wilders’ latest announcement sparked a backlash among Muslims and religious freedom activists worldwide, which soon morphed into a hashtag campaign led by Pakistani Twitter users under:
The hashtag campaign started on July 6 and to date generated more than 481,000 mentions by 128,000 users. On average, each user generated 3.7 mentions of the hashtag, which is a high figure. In scans of organic traffic which @DFRLab conducted previously, a typical average rate of posting ranges from 1.1 to 2.2 posts per user. This suggests some level of artificial automation.
The majority of the users participating in the campaign appear to be from Pakistan; however, a look at the most active accounts revealed that some of those are far more likely to be bots than real Pakistani Twitter users.
A list of 20 most active users who used the hashtag was dominated by new accounts created between July 2, 2018 and July 8, 2018. Most accounts tweeted about nothing else but the hashtag in question. Between them, they generated 11,354 mentions of the hashtag in less than six days, or an average of over 560 posts each, a highly suspicious rate of activity.
Of the 20 most active accounts, 12 accounts were serial retweeters: upward of 98 percent of their Twitter activity consisted of retweets, rather than authored posts.
It was unclear if these users are fully or partially automated, or run by hyperactive account holders. What was clear was that their behavior pattern was conducive with a behavior of a bot, rather than a genuine user.
Apart from hyperactive/bot-like accounts, @DFRLab also observed that hundreds of authored posts using the hashtag were identical copies of one another. This further suggests the use of automated social media behavior to promote the hashtag.
Same Tweets, Different Accounts
One phrase was tweeted 80 times between 7th and 10th of July. The phrase read:
Hazrat Muhammad (PBUH) is the blessing of Allah Almighty for entire universe. Islam preaches to respect all religions. Christianity, Judaism, Buddhism and other religions do not allow violence and hurting others faith. We all must condemn and #Stopblasphemouscartooncontest #Peace
It was launched (and archived here) by user Paras Jahanzaib (@parasjahanzaib1), a news anchor for Pakistani station Sama’a with over 9,000 followers. Her tweet was retweeted over 1,000 times, as well as copied in other tweets word for word.
Similarly, another phrase was used 61 times between July 7 and July 10. The phrase read:
Kindly join the trend Nothings is more important than our Prophet Hazrat Muhammad saw.
Some accounts which tweeted the phrase tagged dozens of other users to join the conversation and use the #Stopblasphemouscartooncontest. This is reminiscent of a common bot strategy @DFRLab observed when 22,000 bots promoted a hashtag and tagged real users in their tweets to get them to use the hashtag.
The following was tweeted by 150 users:
This is not freedom of speech… Its hate speech.!!! The netherland government should stop the contest. My Prophet Muhammad (SAW) is my honour. We muslims will not tolerate anything against him.
It appears to have been launched (and archived here) by user @ahliya_amir, an account which was created in March 2018 and has only 73 followers. It scored just 12 retweets, but garnered more than ten times that number of identical copies.
The spread of another phrase was replicated only 24 times. The phrase read:
Sham on those All who know about this evil activities and they don’t take any action against this blasphemous contest @asmashirazi @ajmaljami @FrontlineKamran @Office_AQPk @Shahidmasooddr @SSEHBAI1 @Intl_Mediatior #StopBlasphemousCartoonContest
The 24 times that it was posted, it had the exact same typo (“sham” instead of “shame”), and tagged the same list of users in the exact same order.
It appeared to have been launched (and archived here) by user @sohailnaik80, an account with only six followers, which made a major effort to amplify the hashtag by asking other high-profile users to use it.
Sham on those All
who know about this evil activities and they don’t take any action against this blasphemous contest @asmashirazi @ajmaljami @FrontlineKamran
— Sohail (@sohailnaik80) July 8, 2018
The more ominous phrase was tweeted by 75 accounts.
#StopBlasphemousCartoonContest We are ready to die No mercy !! Our blood our soul our life Our family Our parents Every thing is ready for sacrifice!
It appeared to have been launched (and archived here) by user @Qurtuba7, an account which was created in May and only has 347 followers, but scored 303 retweets. Such a pattern is not unheard-of in tense political issues such as this, but often indicates artificial amplification.
Overall, there were fewer copies of tweets sharing threatening messages than there were copies of tweets asking people to join the hashtag movement or condemn the cartoon contest. However, there were 14 copies of a tweet saying that the only penalty for blasphemy is beheading on the spot.
Similarly, there were only six copies of a tweet which said:
Pakistan is almost ready to Nuke Netherlands if Geert Wilders is not stopped #NukeNetherlands #StopBlasphermousCartoonContest.
Even the Pope’s Twitter account was targeted by some of these repetitive messages. 12 users tweeted at the Pope’s account and said:
Dear pope, If someone tries to do a blasphemous act against Jesus Christ, U will find us (Muslims) on your back against it. So pls join us to stop a blasphemous cartoon contest #StopBlasphemousCartoonContest.
It is important to note that the accounts tweeting out identical messages listed above did not appear to be a part of a bot network. These accounts were created on different dates, had distinctly different handles and profile pictures. There were no observable patterns between their Twitter timelines, nor the content they amplified. They all, however, used the hashtag #StopBlasphemousCartoonContest extensively. This means that the accounts were most likely run by human users, which points to an organized campaign.
The use of automation and tweet replication indicated that the #StopBlasphemousCartoonContest campaign should be taken with a grain of salt. Nevertheless, @DFRLab found most accounts and tweets using the hashtag to be authentic. This goes to show that not all campaigns using automated amplification and message replication are inauthentic, but it cannot be overlooked that parts of them are.
Ben Nimmo is Senior Fellow for Information Defense at @DFRLab.
Follow along for more in-depth analysis from our #DigitalSherlocks.