Saudi Arabia’s illicit infiltration of Twitter turns out to be only the tip of the iceberg of regional authoritarians’ efforts to wrest control of political discourse on social media.
In early November 2019, the US Department of Justice charged two American residents, one of them a Saudi citizen, with spying on behalf of Saudi Arabia—the first allegation of its kind by the United States against the kingdom. Their target was neither secret US government documents nor military plans, nor industrial blueprints. Instead, the two former employees of the San Francisco-based social media platform Twitter are alleged to have mined the company’s client database for personal information about outspoken Saudi critics of the Saudi regime, and thousands of other Saudi Twitter users.
That the Saudi monarchy would go to such lengths to infiltrate Twitter is indicates the degree to which Twitter, Facebook and other social media platforms—far outgrowing their quotidian origins as casual social chat rooms—have become the de facto modern agora of the Middle East. In a region where public spaces are closely monitored and news media outlets are generally harnessed to the state, independent platforms like Twitter have become the primary public fora where regimes are criticized; where social forces cultivate support and gauge public opinion; and where citizens are ultimately mobilized and coordinated to step into the streets and participate in contentious political action. Such a politically salient space cannot but be the target of manipulation by domestic and foreign actors wishing to shape domestic and regional trajectories.
Indeed, Saudi Arabia’s illicit infiltration of Twitter turns out to be only the tip of the iceberg of the regime’s efforts to wrest control of political discourse on social media. As the kingdom’s ambitious young ruler, Crown Prince Mohammed Bin Salman, insists on retaining and further consolidating monarchic control even while pursuing seismic domestic reforms and embroiling his country in foreign conflicts, he has been proactive about managing the discourse surrounding these activities, both by squelching dissent and putting a positive spin on events. Twitter, as the kingdom’s only plausibly free forum for political debate, must therefore be brought to heel—by manipulating the menu of salient topics that is discussed; by deploying bot armies to disrupt opposition threads and parrot regime positions; and by intimidating, co-opting, arresting and even assassinating oppositional social media “influencers.”
From Liberation Technology to Repressive Instrument
As recently as 2012, social media was heralded as a “liberation technology” that would facilitate a wave of democratization across the Middle East. Indeed, the 2011 Arab uprisings are often remembered as the “Facebook revolutions” or “Twitter revolutions” due to the prominent role social media platforms played as tools for coordinating and mobilizing protestors. But amidst growing political polarization fueled by online manipulation in Western democracies—in the wake of the Cambridge Analytica scandal and after various mass shootings in which social media appears to have played a nefarious role—this earlier euphoria around social media seems absurdly optimistic and sensationalist.
Yet when compared to the world before Twitter, Facebook or YouTube, the advent of social media was nevertheless a profound technological innovation that radically improved the coordination and mobilization of many social movements. Anyone who had access to a computer or smartphone could open a social media account for free. After deactivating geolocational tracking (almost all Middle Eastern users do this) and limiting personally-identifying information, one could post political commentary or calls to action with relative anonymity and impunity, and those posts would be visible to all other users on the platform.
The petro-monarchies of the Gulf, with their enormous resource advantages over civil society, have been learning to take advantage of social media for their own purposes, countering and drowning out dissent at home while shaping narratives around regional affairs.
Contemporary authoritarian regimes are responding to this threatening innovation through a variety of means, but with different capacities. In China, social media platforms like Weibo are managed in-country, so the government can compel the company (Sina) to impose server-side censorship of political tweets and forward the identity of dissident users to state security. Middle Eastern authoritarians, however, must contend with extra-regional platforms like Twitter, whose decisions about content moderation are, at least in theory, beyond their power to affect. On such platforms, civil society activists in the Middle East can incite anti-regime sentiment with text, photos and videos; and can call for protests, name the date and time and location of the next action and agree on an array of tactics—all without the state being able to censor or identify them. In a region where discussing politics in the workplace, coffee shop or even the home is a hazardous undertaking, social media offers a remarkably anarchic space.
But two can play at this game. Regional regimes, particularly the petro-monarchies of the Gulf, with their enormous resource advantages over civil society, have been learning to take advantage of social media for their own purposes by countering and drowning out dissent at home while shaping narratives around regional affairs. With large numbers of their own populations on Twitter (an estimated 9.9 million users in Saudi Arabia), and amidst its significant usership rates elsewhere, the Saudi regime, for example, has been targeting Twitter by using its hashtag vernacular, its borderlessness and its asymmetric landscape of influence in an effort to turn the tables. In recent years, under the direction of Saud al-Qahtani—the former royal court adviser known by activists as the “minister of flies”—the regime has marshalled thousands of social media accounts posing as ordinary citizens, flooding the forum with propaganda and disinformation even while outlawing dissent. At the same time, they have sought to identify, surveil and pressure influential users to voice support for the regime.
Hashtag Gaming and Hijacking
Twitter famously delimits topics of conversation by the hashtag (#) symbol, and users are encouraged to reach relevant audiences by hashtagging their tweets. During the 2011 uprisings, for example, protesters cohered around several famous hashtags, such as #الشعب_يريد_إسقاط_النظام (The People Want the Downfall of the Regime) or #ميدان_التحرير (Tahrir Square). When hashtags show signs of high or rapidly increasing engagement, they are classified (algorithmically) by Twitter as “trending,” where they gain greater visibility and draw greater engagement from a wider audience. Indeed, both of these famous hashtags trended once more in Egypt in September 2019, and in Iraq in October, as Iraq’s 2019 uprising centered on Baghdad’s own Tahrir Square. As part of the broader 2019 uprisings, between October and November, #عراق_ينتفض (Iraq Rises Up) and #لبنان_ينتفض (Lebanon Rises Up) trended in Iraq and Lebanon, respectively, drawing millions of tweets from hundreds of thousands of users.
In a region where discussing politics in the workplace, coffee shop or even the home is a hazardous undertaking, social media offers a remarkably anarchic space.
Arguably, a country’s trending hashtags constitute at first approximation the menu of topics that are “on its mind,” so to speak. Citizens infer what other citizens are talking about (care about, think about) by checking what is trending. Journalists and news stations, though they retain the power to ignore social media and set the agenda themselves via their TV and radio channels and websites, often choose to scour social media trend lists to anticipate the next breaking story. In the kind of self-fulfilling cycle so quintessential to social media, notable events reliably tend to make the trending list, reinforcing our impression that they are notable.
This menu, however, can be manipulated. Given command of enough user accounts, a determined actor can post hundreds or thousands of tweets mentioning a particular hashtag, causing it to trend.
One prominent example of this type of hashtag gaming documented by the researcher Marc Owen Jones took place at the start of the diplomatic crisis between Qatar and the Gulf Cooperation Council members, led by Saudi Arabia. After deploying a military blockade around Qatar, the Saudi government issued an ultimatum that included a list of demands that Qatar should satisfy, including shutting down the Doha-based Al Jazeera network. Soon thereafter, the hashtag #نطالب_بإغلاق_قناة_الخنزيرة (We Demand the Closing of the Channel of Pigs) appeared—a clever use of Arabic wordplay to denigrate the news channel. Analysis of user metadata, however, reveals statistical anomalies indicating that some 70 percent of participating users were centrally commanded accounts (either automated “bots” or human-operated “sockpuppets”) and, moreover, these users dominated the initial volleys of tweets, successfully making the hashtag trend. Many of these accounts, Jones discovered, were created in the months and weeks leading up to the crisis, as if premeditated.
Nor was this an isolated incident. Later that summer, the hashtag #إرحل_يا_تميم (Get Out, Tamim!) trended in Qatar. Ostensibly a domestic cry of discontent against their incumbent Emir, Sheikh Tamim bin Hamad Al Thani, the hashtag turned out on closer inspection to be promoted by scores of bots, and led by several major Saudi and Emirati influencers, including Saudi Arabia’s notorious propaganda chief, Saud Al-Qahtani. The hashtag was quickly countered by pro-Qatari hashtags #تميم_في_كل_مكان (Tamim is Everywhere) and #لا_مشاركة_في_حشتاج_مشبوهة (No Participation in a Suspicious Hashtag) led by major Qatari influencers, including Qahtani’s opposite number, Abdullah Al-Athbah.
At the same time, with thousands of accounts at their command, not only can Gulf regimes promote hashtags consistent with their political positions, but they can also infiltrate those of the opposition. As early as 2016, Jones found that a hashtag favored by human rights defenders in Bahrain was suddenly flooded with anti-Shi’i, sectarian vitriol by hundreds of suspicious accounts. This finding echoes a similar finding in Mexico, where an anti-police hashtag was flooded with tweets containing meaningless strings of symbols, repeatedly forcing protesters to migrate to other hashtags to continue their conversation. These incidents belong to a more widely observed phenomenon of hashtag hijacking, where opportunistic actors, recognizing the heightened visibility of a trending hashtag, seize the moment to convey a message that is off-topic or misrepresents the hashtag’s initial character or intention.
Virtual Foreign Interventions
Twitter conversations tend to be delimited by hashtags, but little else—they have no other borders, which makes them potentially transnational. In the wake of the Gulf crisis with Qatar, and as the Iran-Saudi regional rivalry has escalated, the Gulf has made its influence palpable across the region. Domestic protest hashtags in Egypt, Lebanon and Iraq in 2019, for example, have all felt the pull of Gulf influence. This September, the most retweeted account on Egypt’s Tahrir Square hashtag was that of Turki Shalhoub, a notable anti-authoritarian voice prominent in Gulf hashtags (the user self-reports their location as Saudi Arabia). On the same hashtag, Al Jazeera reporter Husam Yahia was among the top five most retweeted accounts.
The influence of outsiders on domestic Twitter hashtags is potentially exacerbated by the differential penetration of Twitter across Middle Eastern countries. In Egypt, Lebanon and Iraq, Facebook—not Twitter—is the more popular platform. Yet Saudi Arabia, the UAE and Qatar, all favor Twitter, and the international community and news media industry tend to look to Twitter more than to Facebook for political commentary. As a result, domestic voices are under-represented on Twitter, leaving the narrative vulnerable to being shaped by outside elements.
This phenomenon is further exacerbated by Internet shutdowns. When protests broke out in Iraq in October, the Iraqi regime shut down the Internet, causing Iraqi hashtags to be taken up by influencers outside of the country. Indeed, in the initial weeks, the top three influencers on #عراق_ينتفض (Iraq Rises Up) were all pro-Saudi: Amjad Taha (@amjadt25), Sattam Al-Saud (@sattam_al_saud), and Abdullah Al-Bander (@a_albander). All three expressed sympathy and solidarity for the protests, favoring the overthrow of the Iranian-aligned government in Baghdad. Taha in particular reappears as the most retweeted account on the Lebanon uprising hashtags, where pro-Saudi accounts are pleased to observe the ire of protesters directed against Iranian-backed Hezbullah. Taha is one of the most retweeted accounts on Iran’s 2019 protest hashtags, which, again exacerbated by Internet shutdowns, appear dominated by anti-regime voices located abroad.
Influencing the Influencers
While marketing itself as a horizontal space where everyone has a voice, Twitter is, in reality, remarkably unequal. In the same way that some hashtags trend while (most) others do not, some users’ tweets “go viral” and are engaged with to a far greater degree than others. Those users also accumulate more followers, meaning that their future tweets enjoy greater reach and engagement from the beginning. The resulting elite of Twitter influencers, like Turki Shalhoub and Amjad Taha, have an outsize effect on what conversations (hashtags) trend, and what range of opinions are represented within those conversations. This is not to say that others are censored from speaking, but simply that, in the wilderness of voices on social media, many speak but few are heard.
The fact that these influencers wield so much power over political discourse is concerning not only for society—which finds itself expending intellectual energy on a menu of topics effectively chosen by a narrow clique of influencers—but also for the influencers themselves, who increasingly find themselves targets of regime surveillance and repression. Indeed, the former Twitter employees accused of spying for Saudi Arabia did not abuse their privileged data access to identify every user critical of the Kingdom; just the most influential ones.
Omar Abdulaziz, a Saudi political dissident influencer (on both Twitter and YouTube), was identified several years ago—in a now-infamous PowerPoint presentation delivered to Saudi officials by McKinsey consultants—as a vocal and widely followed critic of the Saudi regime. In the summer of 2018, Abdulaziz, who lives in self-imposed exile in Canada, discussed over the phone with Saudi dissident journalist Jamal Khashoggi, then residing in the United States, the idea of building an army of Twitter “bees” to combat Saudi Arabia’s “flies.” Unbeknownst to them, Abdulaziz’s phone had been hacked by advanced Israeli spyware deployed by the Saudi regime. In a recent article, Abdulaziz claims that as many as thirty other anti-regime influencers have confided in him that they have been blackmailed—with material gained by such spyware—to tweet supportively of the regime.
If the threads of conversation are monopolized by so few, regimes can meaningfully steer the discourse by targeting those few with co-optation, intimidation or even assassination. Ironically, over October and November, 2018 in the weeks following the assassination of Jamal Khashoggi, the primary Arabic hashtag concerning Khashoggi exhibited precisely the kind of hyper-concentration of influence that made Khashoggi a target in the first place. 92 percent of tweets on the hashtag were simply retweets of what others said. One pro-Saudi voice, @monther72, single-handedly garnered 8 percent of all retweets, and the top fifty users garnered 53 percent of retweets—all this on a hashtag that involved over 365 thousand users.
Tip of the Iceberg
Seen alongside these other documented tactics, the Saudi Twitter spying scandal appears to be merely the latest attempt, and hardly the last, by a regime determined to contest political discourse on a platform to which it cannot otherwise deny or control access. Across a region where free fora for public debate are largely shut down, platforms like Twitter constitute the last plausible outlet for open expression of political dissent—the de facto agora of the modern Middle East. “Twitter for us is like a parliament,” as one Saudi citizen phrased it.
But in the absence of a real parliament, or indeed any other peaceful, reliable channel by which citizens’ grievances can be heard and addressed, political discontent fomented on Twitter in the Middle East simply has no outlet other than contentious political action, and the violence and unrest that invariably attend it. Such a potent cauldron of discontent cannot be left to bubble and brew on its own. Gulf states, chief among them Saudi Arabia, now intervene regularly—to encourage discontent whenever it threatens to destabilize, say, an Iranian-aligned regime—while confusing and pacifying social movements that threaten authoritarians aligned with the Saudi-Emirati axis.
Whether by influencing their citizens’ thought diets by shaping the menu of topics that trend; or by deploying armies of bots or sockpuppets to parrot regime positions and poison opposition narratives; or by pressuring influential users to toe the party line; the Saudi regime—among others—has meaningfully blunted the liberation potential of social media, and threatens now to turn the tool to its own advantage. And though a country like China, with its top-down, server-side control of platforms, will always be the Orwellian paragon of digital thought authoritarianism, one cannot help but admire the obscene, arthropodic tenacity of feebler dictatorships, such as those of the Gulf, who though lacking the means to directly control platforms, nevertheless learn to adapt to their existence, manipulating, cajoling and gaming them into serving their nefarious ends.
 Marc Owen Jones, “The Gulf Information War| Propaganda, Fake News, and Fake Trends: The Weaponization of Twitter Bots in the Gulf Crisis,” International Journal of Communication 13 (2019), p. 27.
 Marc Owen Jones, “Automated Sectarianism and Pro-Saudi Propaganda on Twitter,” Exposing the Invisible (blog): https://exposingtheinvisible.org/resources/automated-sectarianism.
 The text of Omar Abdulaziz’s complaint against McKinsey can be accessed at: https://digitalcommons.law.scu.edu/cgi/viewcontent.cgi?article=3058&context=historical
 Loveday Morris and Zakaria Zakaria, “Secret recordings give insight into Saudi attempt to silence critics,” Washington Post, October 17, 2018.
 Bill Marczak, John Scott-Railton, Adam Senft, Bahr Abdul Razzak, Ron Deibert “The Kingdom Came to Canada: How Saudi-Linked Digital Espionage. Reached Canadian Soil,” Citizen Lab Research Report No. 115, October 1, 2018: https://citizenlab.ca/2018/10/the-kingdom-came-to-canada-how-saudi-linked-digital-espionage-reached-canadian-soil/.
 Editorial Board, “In Saudi Arabia, Twitter Becomes More Useful to the Repressors than the Repressed,” Washington Post, November 16, 2019.
 Robert F. Worth, “Twitter Gives Saudi Arabia a Revolution of Its Own,” New York Times, October 20, 2012.