The jihadi hunters
A new generation of self-made experts is tracking extremists through their online activity — and rewriting the rules of intelligence in the process
The rise of the Islamic State, also known as ISIS, took much of the world by surprise. When it swept into Mosul and swiftly turned most of northern Iraq into the cornerstone of a regressive new caliphate, the organization was an unknown quantity even to many professional analysts, reporters, and policy makers.
But very quickly, some new go-to sources emerged. Two of them were Twitter streams that unleashed a torrent of crucial links and information. They revealed the depth of the group’s beef with Al Qaeda, which ISIS seemed to consider a higher-priority enemy than even the unbelievers it had executed. They published extracts of the recruitment literature the group had used to lure Western fighters, and shared some of its previously unknown ideological treatises. They brought to light the extensive ISIS propaganda network, while countering some of its claims. Since the United States declared war on the group and started bombing sites in Iraq and Syria, the sources have continued their indispensable work, providing details on little known targets like the “Khorasan Group” and the reaction of ISIS to the American strikes.
These gushers of highly useful information were not coming from inside a formal intelligence operation, or even from the Middle East. Instead, they were being run by ordinary American civilians out of their own homes. One was J.M. Berger, 47, a former journalist turned freelance social network analyst and extremism expert, who published scoop after scoop from his home office in Cambridge. The other was Aaron Zelin, a 26-year-old graduate student in Washington, D.C., who made his name with a blog called Jihadology. The two researchers had been mining the jihadi Internet for years, tracking it with a combination of old-school scholarship and new purpose-built apps.
Zelin and Berger are something new in the intelligence world: part of an emerging breed of online jihadi-hunters who have done pathbreaking work, often independently of government and big media outlets, on a shoestring budget. Numbering less than a dozen, they have earned their reputations over the past four years by being the first to report key developments later confirmed by mainstream research and reporting—such as the split between the Islamic State and Al Qaeda, the burst of jihadi recruitment in the West, and the entry of Hezbollah into the Syrian battle. The meteoric rise of ISIS has been a catalyzing moment for these analysts, pushing them into the spotlight as one of the most important sources of information and context.
These freelance online analysts offer a counterweight to decentralized militant groups. Since the Sept. 11 attacks, America has struggled to grapple with nimble, stateless groups that can move faster than national governments. But the same tools that militant groups and jihadis have exploited so effectively cut both ways. Those who want to shut down violent networks have a new weapon in intelligence-gatherers who operate outside traditional channels and aren’t hindered by bureaucratic myopia.
Despite some friction, their research is forcing the academic and intelligence establishment to treat Twitter, Facebook, and other social media as important sources of data. The small world of social media analysts who have established reliable reputations over time, relying on information freely available in the public domain, implicitly challenges the US government’s claim that only massive, secret surveillance can penetrate jihadi networks.
“Some people say, ‘Who is this guy to be writing about this stuff?’” says Clint Watts, a former FBI agent who has developed an influential following for the analysis of global jihad that he writes after hours. Through Twitter, he’s been able to team up with dozens of experts who 15 years ago he wouldn’t have known how to contact. And through his blog, he’s found a high-level audience that government intelligence analysts could only dream of. “This is mostly a hobby for me,” he said. “The less involved I am in the terrorism analysis community, the more my posts get read.”
Eons ago in Internet time, back in 2006, Hezbollah and Israel fought a quick and destructive war in Lebanon. I was one of the journalists who covered it on the front lines, and we struggled to report the precise nature of Hezbollah’s involvement. Hezbollah tried to obscure its hand, hiding the number of fighters under its command and even whether they were active in the war. It was rare to see a fighter in person at all, and those who were spotted often pretended to represent a local clan or a militia other than Hezbollah.
People guessed at the number of Hezbollah fighters killed, and ultimately had to rely on an unverifiable number issued by the party itself. It was the kind of information that watchers of the conflict had to resign themselves to never being able to know for sure.
During the recent conflict in Syria, Hezbollah denied taking part in combat altogether. But a 27-year-old self-taught analyst named Phillip Smyth, staying up all night in his Washington, D.C., home, began systematically to expose its denial as a lie. Smyth tracked deaths and funerals among Hezbollah supporters; he would identify the same funeral poster on as many as a hundred Facebook pages, then on Hezbollah’s television channel Al-Manar; finally, in some cases, he would telephone friends in Lebanon and ask them to look for and photograph the same poster on a wall. While Hezbollah was still claiming it had no military role in Syria’s civil war, Smyth had proved the group’s leaders were deploying fighters to the Syrian frontlines. He compiled lists of Hezbollah fighters killed in Syrian battles, complete with names and photographs, and posted them in a new Jihadology feature called Hizballah Cavalcade. Months later, Hezbollah finally admitted involvement.
A story like Smyth’s illustrates just how stark a change has come across the once-staid world of intelligence analysts. As recently as a decade ago, this kind of expertise resided almost entirely in government agencies like the CIA and the State Department, or in universities and think tanks with the resources to gather and sift through the data. And it might never surface in public at all.
But today some of the best data are in reach of anyone with an Internet connection—and the Web offers a public platform for anyone able and willing to do the work. Smyth, for example, never worked for the government and didn’t even finish college. He bounced around during adolescence and dropped out of Suffolk University after a year. From an early age he developed a fascination with Lebanon, however, and his mother helped him travel there as a teenager. Smyth learned Arabic and immersed himself in Lebanese culture, obsessively studying the Christian and then the Shia Muslims militant groups that took shape during the civil war. “I was a strange child,” he says.
What started as a lark led him to a job as a researcher at the University of Maryland’s Laboratory for Computational Cultural Dynamics, but his passion was tracking the world of Shia militants. He chatted with them in forums, and built an enormous database of their tweets, Facebook posts, and websites. He listened to their pop music.
As the Arab Spring spiraled into regional upheaval, Smyth would stay up most of the night soaking up the militant movement’s social media feeds. Lurking on Shia militant forums, he learned that rebellion was brewing in Bahrain months before it escalated. Through his careful reading of religious pop music lyrics he learned that Iraqi Shia militias were preparing to join the Syrian civil war. In each case he amassed reams of evidence, running them by Zelin, an acquaintance who ultimately became a close friend. Finally he wrote about his conclusions. Mainstream media, and eventually the US government, picked up his evidence that an internal Syrian civil war had fully morphed into a regional conflagration.
In a sense, Smyth’s work, like Zelin’s and Berger’s, is as new as the media they use: It lies at the intersection of journalism, policy analysis, and intelligence. And fittingly, its practitioners have backgrounds that span those fields. Before he found a perch at a Washington think tank, Zelin was a master’s student with a prolific Web presence. Berger was a freelance journalist before he developed his own apps to scrape and analyze the jihadi Web after a collaboration with Google Ideas.
Clint Watts, on the other hand, was an insider: He spent more than a decade in the military and the FBI before setting out on his own as a consultant with an initially small blog called “Selected Wisdom.” These days, he still prepares his analysis using the same template he learned in 1992 in an entry-level military course, “Intelligence Preparation of the Battlefield.” But it was only after Watts left the government that he began to build his reputation. The amount of information freely available on the Web, he said, was dizzying compared to what he could access on classified government computers. Liberated from the restrictions of classified computer systems and bosses looking for analysis on narrow subjects, he began to follow Al Qaeda and its offshoots across the world. He sparred on Twitter with Omar Hamami, the American who fought with Al Shabab in Somalia. He followed dozens of academic experts, drafting some of them as collaborators in his research.
“I was explaining to my parents just recently that most of what I do would be impossible just 10 years ago,” he said.
If the online ecosystem that these researchers are mining is surprisingly open, it is also uncharted territory. The Islamic State and Hezbollah are both savvy users of online propaganda, which means disinformation as well as the real stuff. For every piece of data Smyth has verified through his research, there are a hundred pieces of misinformation: fake websites, made-up militia names, descriptions of bombings that never happened, and fabricated death announcements. “They try to trick you,” Smyth said. “You’re dealing with some understandably very paranoid people.”
To combat this, the most respected new Web analysts (a group that also includes Charles Lister, a Brookings Institution fellow, and Aron Lund, who edits a Syria blog at the Carnegie Endowment for International Peace) tend to vet each other’s work, and prioritize accuracy over speed. They help each other with translations and argue over interpretation. They collaborate with an ease that traditional analysts, cut off in organizational silos, would scarcely recognize.
Some Twitter analysts do compete to break news first, but much of this crew would rather be late than wrong. Smyth sat for months on evidence of brewing militancy among Bahraini Shia until he could confirm it. In another instance, he encountered a trove of online evidence that Iraqi cleric Moqtada al-Sadr was working with Iran to send fighters to Syria, which would have been explosive news at the time; after months of work he concluded that all of it was fabricated, probably the work of Iranian psyops.
The new breed of online analysts has arisen entirely in the years since Sept. 11, when our government was so notably caught off guard by an underground terrorist threat. The value and speed of their work carries the strong implication that the business needs to change.
That’s not a message the establishment always welcomes. Smyth describes one encounter when a State Department employee bought him a drink to discuss Smyth’s work, before telling him he never could take seriously any research that cites Facebook. Watts recalls the disappointment of another speaker at a Washington conference when she learned that he was reaching all his conclusions without drawing on top-secret sources.
“There are still people who don’t view this as a real form of study,” said Zelin.
But as the online jihadi hunters have risen in prominence, the establishment has increasingly started to embrace their work. White House officials have privately circulated Watts’s memos on the trajectory of global jihad. Berger has found himself in demand as a consultant and commentator, and just got a book deal, his second, to write about ISIS with the Harvard-trained terrorism expert Jessica Stern. Zelin was hired full time as a fellow at the Washington Institute for Near East Policy, and has started a PhD at King’s College London. The university computer lab where Smyth works has asked him to continue his off-duty Hezbollah analysis at the office.
It is easy to imagine that this freewheeling analyst community is a new face of intelligence: decentralized, exciting, and hard to verify, like much about the online world. But it might also be the harbinger of a larger shift. Even in the Twitterverse—the Wild West of people who tweet about terrorism, jihad, and the American policy response—the dozen or so serious analysts in this group are exceptions; it’s much easier to find trolls taunting terrorists or self-appointed experts falling for friendly gambits from propagandists. In interviews, they admit that what they’re doing doesn’t follow an easily replicable template, and they don’t imagine doing it indefinitely. Smyth said it’s exhausting to check his social media sources throughout the night, and even now Zelin said he’s given up on his original goal of being in constant connection with the Twitterstream around the clock.
What they do hope, though, is that their efforts will give a push to the bigger community of well-funded, trained analysts who still provide the bulk of intelligence to the United States. Warfare and communication have changed, and Zelin and his peers expect that ultimately, their work will force academics and the intelligence community to expand their horizons. They will have to accept the value of new kinds of primary-source data, like Twitter, and take seriously the threats and ideas that percolate there.
International jihadis might one day abandon Twitter as quickly as they took it up, and groups like Hezbollah might find ways to police their members on Facebook. But after Facebook and Twitter, more platforms will surely follow. The true lesson of the independent jihadi trackers might be that the intelligence and policy establishment needs to be quicker to follow the culture wherever it chooses to communicate, sometimes leaving secret insights scattered in plain sight.