3

The Truth Is Out There

Social Media and the End of Secrets

 

For nothing is secret, that shall not be made manifest;

neither any thing hid, that shall not be known . . .

—LUKE 8:17

 

 

OPERATION NEPTUNE SPEAR, the mission to get Osama bin Laden, was among the most secretive military missions in history. As the Navy SEAL team took off in the early morning of May 2, 2011, only a few dozen people around the world had been briefed on the operation. One group was clustered in a top-secret military tactical operations center, and the other was gathered around a table in the White House Situation Room. There, President Obama and his advisors tracked the SEALs’ progress from 7,600 miles away via a direct video link that was the sole source of information about the mission.

Or at least it was supposed to be. No one had counted on @ReallyVirtual.

@ReallyVirtual wasn’t a spy. He wasn’t a journalist either. His real name was Sohaib Athar, a Pakistani tech geek and café owner, whose social media handle described him as “an IT consultant taking a break from the rat-race by hiding in the mountains with his laptops.”

A few years earlier, Athar had moved from the busy city of Lahore to the more pleasant town of Abbottabad, a mountain tourist hub and home of the Pakistan Military Academy—as well as, now, the most wanted man in the world. Crashing on a late-night software project, Athar was distracted by the sound of helicopters overhead. So he did what millions of people do each day: he took to social media to complain.

“Helicopter hovering above Abbottabad at 1AM (is a rare event),” he tweeted first. As the SEALs’ mission played out over the next several minutes, Athar posted a litany of complaints that doubled as news reports. When the first helicopter took off, carrying away bin Laden’s body and hard drives filled with data on Al Qaeda’s networks, he tweeted, “Go away helicopter—before I take out my giant swatter :-/.” As the remaining SEALs detonated a crashed helicopter and piled into a backup chopper, Athar shared the news of the explosion. “A huge window shaking bang here in Abbottabad . . . ,” he tweeted. “I hope its not the start of something nasty :-S.”

Eight hours later, the traditional news media finally caught up to one of the most important stories in a decade. On NBC, The Celebrity Apprentice was airing. Just as Donald Trump was explaining his rationale for “rehiring” singer La Toya Jackson, the network cut away. In a surprise prime-time address, President Obama announced that a top-secret raid had taken place in Pakistan and that Osama bin Laden was dead. “Justice has been done,” he concluded. In cities around the United States, people danced in the streets. Thousands of miles away, Athar was coming to his own realization. “Uh oh,” he tweeted, “now I’m the guy who liveblogged the Osama raid without knowing it.”

It was lunchtime in Abbottabad when the messages began pouring in—a trickle that swiftly transformed into a torrent. Athar’s Twitter follower count jumped from 750 to 86,000. He was deluged with calls for interviews and fan requests. Local journalists sped to his café to talk face-to-face. More worrying, a growing online mob accused him of spying for either the U.S. or Pakistani government. Surely, they argued, that was the only way that Athar could have known about such a top-secret military operation.

But the truth was simpler and more profound: Sohaib Athar was just a guy who happened to be near something newsworthy, with a computer and a social media account at hand.

When the internet first began to boom in the 1990s, internet theorists proclaimed that the networked world would lead to a wave of what they called “disintermediation.” They described how, by removing the need for “in-between” services, the internet would disrupt all sorts of longstanding industries. Disintermediation soon remade realms ranging from retail stores (courtesy of Amazon) and taxi companies (courtesy of Uber) to dating (courtesy of Tinder). Athar’s tale showed how the business of information gathering had undergone the same kind of disintermediation. No longer did a reporter need to be a credentialed journalist working for a major news organization. A reporter could be anyone who was in the right place at the right time. But this shift wasn’t just about reporting the news. It was also changing all the people who make use of this information, be they citizens, politicians, soldiers, or spies.

There was one more lesson wrapped up in the surreal saga of Sohaib Athar, unnoticed by many at the time but painfully obvious to observers in the U.S. intelligence community. Operation Neptune Spear—among the most closely guarded operations in history—had nonetheless been documented in real time for anyone in the world to see. And this had happened accidentally, in a country where just 6 percent of the population had internet access at the time. What would the future hold as more and more people came online? Even more, how would intelligence agencies cope when it wasn’t just a lone night owl inadvertently sharing secrets, but organized groups of analysts dedicated to parsing social media to find the operations hidden in plain sight?

“Secrets now come with a half-life,” one CIA official told us, with more than a twinge of regret.

EVERYTHING IS ILLUMINATED

“Let’s give a welcome to ‘Macaca’ here,” puffed the fleshy man in the blue shirt, pointing toward the camera. “Welcome to America and the real world of Virginia.”

It was August 2006, and Senator George Allen was barnstorming for votes at a rural Virginia park. A darling of the conservative wing of the Republican Party, he was already looking past this reelection bid. He’d recently made exploratory trips to Iowa and New Hampshire, testing the waters before a potential presidential run. But although Allen didn’t know it yet, his political career had just ended in that single moment, all because of how the internet had changed.

The man behind the camera was S. R. Sidarth, a 20-year-old volunteer for Allen’s opponent, who had taken to filming Allen’s events. Sidarth was also Indian American—and the only brown face amid the rally’s 100 attendees. And the name Allen had just called him, “Macaca,” is Portuguese for “monkey,” used as a racial slur for centuries.

The history of politicians saying and doing horrible or stupid things on the campaign trail is as old as democracy itself. But the trajectory from a bad moment to a fatal gaffe previously required a professional journalist to be on the scene to document it. Then the moment would have to be reported via a newspaper or radio or TV station. In order for the gaffe to build truly national momentum, other professional journalists and their outlets would then have to pick it up. Unfortunately for Allen, social media had altered this process, propelling his words beyond the control of any politician or journalist.

Sidarth’s minute-long recording was quickly posted on YouTube, the new video-sharing platform, scarcely a year old in 2006. This was an unusual decision at the time, because the video clip was unedited and unattached to any broader story. It proved an ingenious move, however, as the very nature of the clip was part of its appeal. Easy to view and share, Sidarth’s video went viral, with hundreds of thousands of people seeing it firsthand online, and the news media being able to report on and link to it.

Allen’s advisors, skilled and experienced in the old model of political campaigning, were flummoxed. At first they denied that the incident had happened. Then they claimed that Allen had done nothing wrong, explaining that “Macaca” wasn’t meant as a slur. And then they pivoted to claiming that Allen had actually said “Mohawk,” referring to Sidarth’s hair.

The problem with each explanation was that, unlike in the past, anyone who wanted to could now see the evidence for themselves. They could click “play” and hear the ugly word again and again. They could see that Allen was using it to describe the one brown-skinned person in a crowd of white people and suggesting that Sidarth wasn’t a “real” American.

Allen’s lead in the polls plummeted, and he went on to lose a race in which his victory had been all but guaranteed. Instead of making a run for president, he never served in elected office again. As for Sidarth, he was named Salon’s person of the year: a “symbol of politics in the 21st century, a brave new world in which any video clip can be broadcast instantly everywhere and any 20-year-old with a camera can change the world.”

What became known as the “Macaca moment” was a hint of the web-driven radical transparency that was just starting to change how information was gathered and shared—even the nature of secrecy itself.

The relatively new digital camera that Sidarth used to document Allen’s fateful words has since been followed by some 9 billion digital devices that, importantly, are now linked online. By 2020, that number will soar to 50 billion, as devices ranging from smartphones to smart cars to smart toothbrushes all join in to feed the internet.

Most significantly, all of the new items coming online carry something that the computers used by ARPANET, and even the one used by Mark Zuckerberg to create Facebook, lacked: “sensors,” devices for gathering information about the world beyond the computer. Some sensors are self-evident, like the camera of a smartphone. Others lurk in the background, like the magnometer and GPS that provide information about direction and location. These billions of internet-enabled devices, each carrying multiple sensors, are on pace to create a world of almost a trillion sensors. And any information put online comes with “metadata,” akin to digital stamps that provide underlying details of the point of origin and movement of any online data. Each tweet posted on Twitter, for instance, carries with it more than sixty-five different elements of metadata.

This plethora of sensors and associated metadata is making real an idea that has long possessed (and frightened) humanity: the possibility of an ever-present watcher. The ancient Greeks imagined it as Argus Panoptes, a mythological giant with 100 eyes. During the Enlightenment, the English philosopher Jeremy Bentham turned the monster into the Panopticon—a hypothetical building in which all the occupants could be observed, but they never saw those watching them. Ominously, Bentham pitched his design as being useful for either a factory or a prison. George Orwell then gave the panoptic idea an even darker spin in his novel 1984. His futuristic totalitarian world was filled with “telescreens,” wall-mounted televisions that watched viewers while they watched the screens.

Today, the combination of mass sensors and social media has rendered these bizarre fantasies an equally bizarre reality. Yet rather than gods or rulers, we are collectively the ones doing the watching. A decade after Allen flamed out, any politician worth their salt, gazing into a crowd of a hundred people, might reasonably assume they are the subject of no less than a half dozen videos and many more photographs, texts, and audio snippets, any of which might prompt scores of social media reactions. Indeed, they would likely be upset if no one posted about the event online. To ensure that didn’t happen, they would likely be doing it themselves. In the run-up to the 2018 U.S. midterm elections, some candidates were pushing out more than a dozen Facebook videos a day.

None of this means that gaffes like Allen’s no longer happen, nor that racist comments are no longer made; rather the opposite. When everything can be recorded, everything is on the record.

The amount of data being gathered about the world around us and then put online is astounding. In a minute, Facebook sees the creation of 500,000 new comments, 293,000 new statuses, and 450,000 new photos; YouTube the uploading of more than 400 hours of video; and Twitter the posting of more than 300,000 tweets. And behind this lies billions more dots of added data and metadata, such as a friend tagging who appeared in that Facebook photo or the system marking what cellphone tower the message was transmitted through. In the United States, the size of this “digital universe” doubles roughly every three years.

Each point of information might be from an observer consciously capturing a speech or a gun battle, or it might be unwittingly shared with the world, as was Athar’s coverage of the bin Laden raid. The most valuable information might even lurk in the background. Snapping tourist photos of a harbor, Chinese civilians once accidentally revealed secrets of their navy’s new aircraft carrier, under construction in the distance. Or an interesting tidbit might lie in the technical background. Exercise apps have inadvertently revealed everything from the movements of a murderer committing his crime to the location of a secret CIA “black site” facility in the Middle East. (A heat map made from tracing agents’ daily jogs around the perimeter of their base provided a near-perfect outline of one installation.)

In 2017, General Mark Milley, chief of staff of the U.S. Army, summed up what this means for the military: “For the first time in human history, it is near impossible to be unobserved.” Consider that in preparation for D-Day in June 1944, the Allies amassed 2 million soldiers and tens of thousands of tanks, cannons, jeeps, trucks, and airplanes in the British Isles. Although German intelligence knew that the Allied forces were there, they never figured out where or when they would strike. That information came only when the first Americans stormed Utah Beach. Today, a single soldier’s or local civilian’s Facebook account would be enough to give away the whole gambit. Indeed, even their digital silence might be enough to give it away, since a gap in the otherwise all-encompassing social media fabric would be conspicuous.

What’s unveiled is not just the movement of armies. Such data can be used to geographically pinpoint people, even in circumstances where they’d rather not be found. For instance, Ashley Madison is a social network that links people thinking of cheating on their spouses. Its algorithms mine social media to detect when business travelers arrive at a hotel (and thus are more likely to stray from their marriages). In a similar fashion, in the fighting that began in Ukraine in 2014, Russian military intelligence pinpointed the smartphones of Ukrainian soldiers arriving on the front lines. Just as Ashley Madison uses geographically targeted data to fire off web ads to potentially philandering travelers, the Russians used it to send messages like “They’ll find your bodies when the snow melts.” Then their artillery began firing at the Ukrainians.

But what stands out about all this information is not just its massive scale and form. It is that most of it is about us, pushed out by us. This all arguably started back in 2006, when Facebook rolled out a design update that included a small text box that asked a simple question: “What’s on your mind?” Since then, the “status update” has allowed people to use social media to share anything and everything about their lives they want to, from musings and geotagged photos to live video feeds and augmented-reality stickers.

The result is that we are now our own worst mythological monsters—not just watchers but chronic over-sharers. We post on everything, from events small (your grocery list) to momentous (the birth of a child, which one of us actually live-tweeted). The exemplar of this is the “selfie,” a picture taken of yourself and shared as widely as possible online. At the current pace, the average American millennial will take around 26,000 selfies in their lifetime. Fighter pilots take selfies during combat missions. Refugees take selfies to celebrate making it to safety. In 2016, one victim of an airplane hijacking scored the ultimate millennial coup: taking a selfie with his hijacker.

These postings are revelatory of our personal experiences, of course. But they also now convey the weightiest issues of public policy. The first sitting world leader to use social media was Canadian prime minister Stephen Harper in 2008, followed quickly by U.S. president Barack Obama. A decade later, the leaders of 178 countries have joined in. Even former Iranian president Mahmoud Ahmadinejad, who famously banned Twitter during a brutal crackdown, has since changed his mind on the morality—and utility—of social media. He debuted online with a friendly English-language video as he stood next to the Iranian flag. “Let’s all love each other,” he tweeted.

It is not just world leaders, though. Agencies at every level and in every type of government now share their own news, from some 4,000 national embassies to the fifth-grade student council of the Upper Greenwood Lake Elementary School. National militaries have gotten in the game as well. In the United States, the Army, Navy, Air Force, and Marine Corps all have an official social media presence. So do their bases, combat units, generals, and admirals. Even the status of individual military operations is now updated. When the U.S. military’s Central Command expanded Operation Inherent Resolve against ISIS in 2016, Twitter users could follow along directly via the hashtag #TalkOIR. A U.S. military officer even popped up on the discussion forum Reddit for its signature Ask Me Anything series. He answered dozens of questions on the state of anti-ISIS operations, but delicately declined to offer the U.S. military’s thoughts on the latest season of the Archer television show.

The result of all this sharing is an immense, endlessly multiplying churn of information and viewpoints. And it matters not just for the here and now. Nothing truly disappears online. Instead, the data builds and builds, just waiting to reemerge at any moment. According to law professor Jeffrey Rosen, the social media revolution has essentially marked “the end of forgetting.”

This massive accumulation of updates, snapshots, and posts over time offers revelations of its own. The clearest exemplar of this phenomenon is the first president to have used social media before running for office. As both a television celebrity and social media addict, Donald Trump entered politics with a vast digital trail behind him. The Internet Archive has a fully perusable, downloadable collection of more than a thousand hours of Trump-related video, while his Twitter account has generated some 40,000 messages. Never has a president shared so much of himself—not just words, but even neuroses and particular psychological tics—for all the world to see. Trump is a man—the most powerful in the world—whose very essence has been imprinted on the internet.

Tom Nichols is a professor at the U.S. Naval War College who worked with the intelligence community during the Cold War. He explains the unprecedented value of this vault of information: “It’s something you never want the enemy to know. And yet it’s all out there . . . It’s also a window into how the President processes information—or how he doesn’t process info he doesn’t like. Solid gold info.” Russian intelligence services reportedly came to the same conclusion, using Trump’s Twitter account as the basis on which to build a psychological profile of the forty-fifth president.

And exhaustive though it is, Trump’s online dossier encompasses only a single decade and started only when he was in his 60s. Nearly every future politician or general or soldier or voter will have a much bigger dataset, from much earlier in life. Indeed, this inescapable record may well change the prospects of those who wish to become leaders in the future. As Barack Obama said after he left office, “If you had pictures of everything I’d done in high school, I probably wouldn’t have been President of the United States.”

The consequences of such extensive online sharing go beyond broadcasting our day-to-day activities and thoughts. Social media can also provide a surprisingly clear window into our psychological and neurological states. Luke Stark, a researcher in the sociology department at Dartmouth College, explains that accumulated online postings provide “something much more akin to medical data or psychiatric data.” Even the most trivial details can be unexpectedly revealing. Consistent use of black-and-white Instagram filters and single-face photos, for instance, has proven a fairly good identifier of clinical depression.

As unprecedented as all this information may be, it matters little unless there’s someone on the other end to appreciate—or exploit—it. Just as the internet has changed the volume, source, and availability of content, it also has wrought dramatic changes in how this information is used.

THE ELECTRIC BRAIN AWAKENS

The ten killers snuck into Mumbai’s port aboard inflatable boats on November 26, 2008. Once ashore, they split up, fading into the megacity of some 18 million people. The attacks started soon after: a staccato of massacres at a railway station, a tourist café, a luxury hotel, and a synagogue. Over the next three days, 164 civilians and police officers would be killed. Another 300 would be injured. The tragedy would mark the deadliest Indian terror attack in a generation. It also signaled a radical change in how the news was both parsed and spread.

Although there were just 6 million Twitter users worldwide at the time, Mumbai’s booming IT sector had helped build a small, vocal network of early adopters. The tweets began minutes after the first attack: 140-character reports, observations, cries for help, and words of warning that captured each explosion and gunshot. “I have just heard 2 more loud blasts around my house in colaba,” wrote @kapilb, blocks from one attack. “Grenades thrown at colaba,” added @romik a few minutes later.

Users living thousands of miles from the attack also served as unexpected conduits for those trapped inside it. When the terrorists took hostages in two Mumbai hotels, a Silicon Valley–based venture capitalist helped spread the news. “I just spoke with my friends at the Taj and Oberoi,” tweeted @skverma, “people have been evacuated or are barracaded in their Rooms.”

With Indian authorities reeling and journalists restricted as Mumbai became an instant and unexpected war zone, much of the useful reporting originated from a network that emerged spontaneously. Mumbai’s online community kicked into gear, sharing riveting stories that quickly spread across the digital ecosystem. One brave resident took to the streets, snapping dozens of pictures. He posted them to the image-sharing service Flickr, originally created for video gamers. In a reversal of journalistic practice, these amateur photographs filled the front pages of newspapers the next day.

Of course, as this network of amateur reporters recounted the Mumbai attacks, it also spread all the spurious rumors that we know now accompany such events. There were fake reports of nonexistent attacks, which then begat more falsehoods as people reacted to them. So new communities arose to sift through the mountain of data, networks of online analysts separating fact from fiction.

In a move that would soon become the norm, the Mumbai attacks got their own Wikipedia page—roughly four hours after the first shots had been fired. Dozens of volunteer editors pitched in, debating everything from serious allegations of external support (rumors of Pakistani government involvement were already swirling) to tricky issues of phrasing (were the attackers “Muslim militants” or “Muslim terrorists”?). All told, before the last terrorist had been cornered and shot, the Wikipedia entry was edited more than 1,800 times.

Another important new tool was put to work. Google Maps, launched in 2005, had made it possible for people scattered across the world to locate and share precise coordinates—a capability previously reserved for only the most advanced militaries. The location of each bomb blast and firefight was plotted as soon as it was reported. This revealed something else unexpected. It became possible not just to track the latest news, but to build the operation’s history. You could trace where the terrorists had disembarked from their boats; where the first car bomb had detonated; where each gun battle had taken place. Eventually, Google Maps would even plot the location of victims’ funerals.

Anyone in the world could now monitor a battle unfolding in real time. This included even the people who had sent the attackers. Nesting in a control room in Karachi, Pakistan, the operation’s commanders reportedly kept in cellphone contact with the gunmen on the ground, guiding them from target to target. Rather than relying on some secret intelligence network for updates, they could track the same social media platforms as everyone else. Just as a tweet about a possible explosion warned others to stay away, it helped the militants predict the focus of attention and the paths of emergency responders.

Eventually, the online crowd began realizing this and urged others to stop talking about the movement of Indian security forces. Some took it upon themselves to police the news. One widely shared message declared, “Indian government asks for live Twitter updates from Mumbai to cease immediately. ALL LIVE UPDATES—PLEASE STOP TWEETING.” The Indian government had said no such thing; it was fake news. Others began to fight back the only way they knew how, sending a flurry of tweets at the terrorists. “Die, die, die, if you’re reading this,” one wrote.

The online activity also begat a new kind of emergency coordination. Messages poured in begging for blood donations, directing donors to the hospital where most of the victims had been taken. Other users spread word of tip lines, publicizing the information further and faster than the Indian government could on its own. This was a radical departure from past emergencies, which had depended on sluggish public broadcast systems and word of mouth.

As the smoke cleared, the Mumbai attack left several legacies. It was a searing tragedy visited upon hundreds of families. It brought two nuclear powers to the brink of war. And it foreshadowed a major technological shift. Hundreds of witnesses—some on-site, some from afar—had generated a volume of information that might previously have taken months of diligent reporting to gather. By stitching these individual accounts together, the online community had woven seemingly disparate bits of data into a cohesive whole. It was like watching the growing synaptic connections of a giant electric brain.

There is a word for this: “crowdsourcing.” An idea that had danced excitedly on the lips of Silicon Valley evangelists for years, crowdsourcing had been originally conceived as a new way to outsource software programming jobs, the internet bringing people together to work collectively, more quickly and cheaply than ever before. As social media use had skyrocketed, the promise of crowdsourcing had extended into a space beyond business. Mumbai proved an early, powerful demonstration of the concept in action. The incidents would swiftly multiply from there.

At its core, crowdsourcing is about redistributing power—vesting the many with a degree of influence once reserved for the few. Sometimes, crowdsourcing might be about raising awareness, other times about money (also known as “crowdfunding”). It can kick-start new businesses or throw support to people who might once have languished in the shadows. It was through crowdsourcing, for instance, that septuagenarian socialist Bernie Sanders became a fundraising juggernaut in the 2016 U.S. presidential election, raking in $218 million online.

Of course, like any useful tool, crowdsourcing has also been bent to the demands of war. A generation ago, Al Qaeda was started by the son of a Saudi billionaire. By the time of the Syrian civil war and the rise of ISIS, the internet was the “preferred arena for fundraising” for terrorism, for the same reasons it has proven so effective for startup companies, nonprofits, and political campaigns. It doesn’t just allow wide geographic reach. It expands the circle of fundraisers, seemingly linking even the smallest donor with their gift target on a personal level. As The Economist explained, this was, in fact, one of the key factors that fueled the years-long Syrian civil war. Fighters sourced needed funds by learning “to crowdfund their war using Instagram, Facebook and YouTube. In exchange for a sense of what the war was really like, the fighters asked for donations via PayPal. In effect, they sold their war online.”

Just as any digital marketing guru would advise, Syrian fighting groups have molded their message to reflect the donor pool’s interest. Many of Syria’s early rebel fighters sought to establish a free, secular democracy. But this prospect didn’t excite the fundamentalist donors from the wealthy Arab states. So, to better sell their effort online, even secular fighters grew impressively long beards and made sure to pepper their battle videos with repetitions of “Allahu Akbar” (God is great).

Fundraisers also got creative through what became known as “financial jihad.” Some clerics argued that online pledges allowed donors to fulfill their religious duties in the same way they would if they had actually served in battle. Much as you might sponsor your cousin’s “fun run” to fight cancer, you could sponsor a rocket-propelled grenade (RPG) launcher for a Syrian rebel (it went for $800). Or you could back the rebels’ opponents. Hezbollah, the Iranian-sponsored terror group that allied with the Syrian regime, ran an “equip a mujahid” campaign on Facebook and Twitter. It similarly allowed online supporters to fulfill their religious obligations by buying weapons and ammunition for the war.

As radical transparency merges with crowdsourcing, the result can wander into the grotesque. In 2016, a hard-line Iraqi militia took to Instagram to brag about capturing a suspected ISIS fighter. The militia then invited its 75,000 online fans to vote on whether to kill or release him. Eager, violent comments rolled in from around the world, including many from the United States. Two hours later, a member of the militia posted a follow-up selfie; the body of the prisoner lay in a pool of blood behind him. The caption read, “Thanks for vote.” In the words of Adam Linehan, a blogger and U.S. Army veteran, this represented a bizarre evolution in warfare: “A guy on the toilet in Omaha, Nebraska, could emerge from the bathroom with the blood of some 18-year-old Syrian on his hands.”

The rapidity with which these bloodthirsty votes came in illustrates how the speed at which these spontaneous online collectives form is accelerating apace with the information environment. In 2008, the online mosaic of the Mumbai terror attacks was pieced together in a matter of hours. Five years later, at the Boston Marathon Bombing of April 15, 2013, this timeline had shifted by an order of magnitude.

It took about thirty seconds for Boston’s emergency coordination center to learn of the attack that killed 3 people and wounded nearly 300 more—a fact that it would proudly document in an after-action report commissioned a year later. But the news was already online. The very moment when Boston police officers and firefighters were shouting into their radios, @KristenSurman tapped out a frantic message to her Twitter followers: “Holy shit! Explosion!” Seconds later, @Boston_to_a_T uploaded the first photo of the attack, taken mid-fireball. It took three minutes for the terror attack to be reported by a professional media outlet. The coverage came via a quick tweet from Fox Sports Radio—but not in Boston; in Washington State. By contrast, it took nearly an hour for Boston police to formally confirm the bombing.

Since then, the size and reach of this online audience has continued to grow exponentially, while the number of global smartphone users has more than doubled. Virtually any event leaves a digital trail that can be captured, shared, and examined by hungry internet users. As the audience eagerly rushes from one development to the next, it drives new developments of its own. Less and less is there a discrete “news cycle.” Now there is only the news, surrounding everyone like the Force in Star Wars, omnipresent and connected to all.

The best way to describe the feeling that results is a term from the field of philosophy: “presentism.” In presentism, the past and future are pinched away, replaced by an incomprehensibly vast now. If you’ve ever found yourself paralyzed as you gaze at a continually updating Twitter feed or Facebook timeline, you know exactly what presentism feels like. Serious reflection on the past is hijacked by the urgency of the current moment; serious planning for the future is derailed by never-ending distraction. Media theorist Douglas Rushkoff has described this as “present shock.” Buffeted by a constant stream of information, many internet users can feel caught in a struggle just to avoid being swept away by the current.

With the marriage of radical transparency and ever faster and more feverish crowdsourcing, the line between observer and participant has been irrevocably blurred. “News” originates not just with journalists, but with anyone at the scene with a smartphone; any soldier with an Instagram account; any president tweeting away while watching Fox News in his bedroom. In a sense, everyone has become part of the news. And while people who serve to make sense of the madness still exist, the character and identity of these gatekeepers have transformed as well.

THE NEW MIDDLE

The word “media” comes from the Latin word for “middle.” For the past century, “media” has been used to refer to the professional journalists and news organizations paid to serve as the conduit between the public and the news. Today, social media has put new voices in the middle.

At the ripe old age of 11, Rene Silva joined this once-august profession. He started his first newspaper, printing it on his school’s photocopier. Soon Silva was running a full-fledged online news agency, dedicated to catching the stories that fell through the cracks. He built a website, he launched a Facebook page and YouTube channel, and he maintained an active Twitter feed. All this before he left high school.

Familiar though his story might sound, Silva had none of the advantages of a typical tech-savvy American teenager from suburbia. Instead, his Voz das Comunidades (Voice of the Community) reported on life in Complexo do Alemão, a cluster of Brazilian favelas infested with drug traffickers, with 70,000 people squeezed into a single square mile. Although most of Rene’s stories resembled those of any local newspaper—articles on illegal parking and profiles of local leaders—sometimes circumstances caught up with him. In 2010, when the Brazilian government’s security forces launched a series of rolling firefights against the gangs, the 17-year-old Silva was there to capture it all. He live-tweeted the location of each battle, set up a video feed, and enlisted friends to help hunt for more stories. At times he took to correcting the accounts of his rivals—professional, adult journalists, who didn’t know the neighborhood and so often misreported the street names. Silva’s efforts won him international acclaim, and he was named an honorary torchbearer in both the 2012 and 2016 Olympic Games. Today, he seeks to apply his hyperlocal reporting approach to neighborhoods across Brazil.

Just as social media has altered the people who witness and report the news, it has also changed those who deliberately go out and gather it. Across the world, there is a new breed of journalist, empowered by the web, often referred to as the “citizen reporter.”

Sometimes, these reporters fill in the gaps, using social media to report on areas the traditional media finds unprofitable to report from. Selinsgrove, Pennsylvania, population 5,500, is far more bucolic than the Alemão favelas, but it, too, lacks a truly local paper. Into the void stepped 9-year-old Hilde Kate Lysiak, whose Orange Street News has covered everything from vandalism to a fire department corruption scandal. Hilde’s journalistic chops are perhaps best illustrated by how she responded when police tried to keep a murder investigation quiet, asking her not to report it. She refused. “I may be nine, but I have learned that my job as a reporter is to get the truth to the people. I work for them, not the police.”

Other times, they cover stories in places where the job of reporting has become too dangerous. For instance, Mexico’s decade-long drug war between the government, paramilitaries, and cartels has taken a terrible toll on society—and journalists. There were nearly 800 documented attacks on reporters between 2006 and 2016, killing dozens.

As they seek to avoid this fate, media organizations face a terrible set of choices. Some have become criminal mouthpieces, employing cartel-affiliated reporters known as “links.” Others have become more selective in the kinds of news they cover. The Norte newspaper in Juárez took the second route, avoiding stories that might anger the state’s secret cartel overseers. As one editor explained, “You do it or you die, and nobody wants to die. Auto Censura—self-censorship—that’s our shield.” But even that wasn’t enough to hold the cartels at bay. In 2017, the Norte shuttered with a final, bitter “Adiós.”

Into this void stepped Felina, a woman whose Twitter handle and image were inspired by Catwoman, of comic book legend. Her home state of Tamaulipas had been split between two gangs, the Gulf Cartel and the Zetas, whose war had killed over 15,000 people. Faced with this ongoing horror, she and a small group of online vigilantes banded together to form an organization to inform and protect their fellow citizens. They called it Valor por Tamaulipas (Courage for Tamaulipas). With Felina serving as administrator, the group built a crowdsourced news service that gathered and distributed the information that residents of Tamaulipas needed to know to stay safe. The reports ranged from notifications of ongoing shootouts to photos of trigger-happy cartel members, so citizens would know which streets and people to avoid.

Felina was soon curating one of the most popular social media channels in the state, with over half a million followers on Facebook and 100,000 on Twitter. This success, though, marked Felina as a target. A cartel offered a reward of 600,000 pesos (about $48,000 at the time) for the identities of the site’s anonymous administrators. It even posted its own message to the citizen journalists’ feed: “We’re coming very close to many of you . . . watch out felina.”

Felina was undeterred. The threats attracted even more attention—the internet’s most powerful currency. Web traffic quadrupled, and fans from near and far celebrated their work. Inspired, Felina threw herself into a grander vision of transforming Tamaulipas. She labored to raise money for poor people, organizing blood donor drives and even helping to find a needy person a pair of orthopedic shoes.

This generosity of spirit may have been Felina’s undoing, seeding clues that the cartels used to track her down. On October 16, 2014, Felina’s Twitter account posted the following message: “FRIENDS AND FAMILY, MY REAL NAME IS MARÍA DEL ROSARIO FUENTES RUBIO. I AM A PHYSICIAN. TODAY MY LIFE HAS COME TO AN END.”

The feed that had long sought to help the citizens of Tamaulipas then posted two photos in rapid succession. The first showed a middle-aged woman looking directly into the camera. The second showed the same woman lying on the floor, a bullet hole in her head. As journalist Jason McGahan put it, “She tweeted against the Mexican cartels. They tweeted her murder.”

With such risks, the work of these brave citizen reporters becomes something more than mere reporting. Perhaps the most notable example unfolded in the midst of the terrible group ISIS, whose rise also personified social media’s new power. In 2013, the city of Raqqa, Syria, fell to ISIS, becoming its capital. Soon, Raqqa became an epicenter of horror, from the brainwashing of children to public crucifixions. Yet, because ISIS had a practice of murdering professional journalists, there was no one from the international media to document and report this reign of terror.

So a group of seventeen citizens banded together to tell the story of their city’s destruction. They did so via an online news network they called Raqqa Is Being Slaughtered Silently. It was as much an act of resistance as reporting. Their belief, as one member put it, was that “truth-telling” would prove to be more powerful than ISIS’s weapons.

For years, these citizen reporters were the primary source of information on life in a city ruled by ISIS. The group’s work was incredibly dangerous, with ISIS on the constant hunt for them and anyone they loved. Soon after the network’s launch, ISIS began broadcasting counterprogramming titled “They Are the Enemy So Beware of Them.” It showed a group, whom ISIS claimed to be the reporters and their families, being paraded in front of cameras and then hanged from a tree. Raqqa Is Being Slaughtered Silently then reported these executions, sadly noting that ISIS had killed the wrong people. Eventually, ten members of the network were found and killed. One woman, Ruqia Hassan, dashed off a quick Facebook post shortly before ISIS police captured and executed her. “It’s okay because they will cut my head,” she wrote. “And I have dignity, it’s better than I live in humiliation.” For both its bravery and novelty, the group was awarded a 2015 International Press Freedom Award.

A common thread runs through all of these stories. From favela life to cartel bloodlettings to civil wars, social media has erased the distinction between citizen, journalist, activist, and resistance fighter. Anyone with an internet connection can move seamlessly between these roles. Often, they can play them all at once.

This communications revolution has been complemented by another shift, easier to miss but perhaps even more consequential. Just as the people who document and reveal secrets to the world have changed, so, too, have the people who traditionally work behind the scenes collecting and analyzing this information. They’re called “intelligence analysts,” or, more colloquially, “spies.” And today, they look (and work) quite a bit different indeed.

WORK-FROM-HOME SHERLOCK HOLMES

As he boarded his long-haul flight from the Netherlands to Malaysia on July 17, 2014, the Dutch musician Cor Pan snapped a picture of the waiting Boeing 777 and uploaded it to Facebook. “In case we go missing, here’s what it looks like,” he wrote. It was supposed to be a joke. Instead, it would become one of the last digital echoes of a terrible tragedy.

A few hours later, the plane was flying over eastern Ukraine, a region divided between the local government and Russian-backed separatists. Many of the 298 passengers and crew aboard were drowsing with the shades drawn. They were a mix of vacationers, business travelers, and a group of scientists bound for an HIV/AIDS conference. In the cockpit, the mood was similarly placid. The pilots’ biggest worry was light turbulence. We know this because all the sounds and activity on the cockpit voice recorder were perfectly normal, up until the thirty milliseconds before it stopped working.

The pilots never saw the missile as it pierced the cloud cover to their left. In a fraction of a second, over 7,600 pieces of superheated shrapnel tore through the cockpit, ripping the pilots to shreds. The blast cleaved the front of the plane from the rest of the fuselage. As the aircraft jerked and began to fall, it separated into three pieces. Many passengers in the cabin remained alive through the plane’s plummet, struggling to understand what was happening. For ninety seconds, they were overwhelmed with deafening sounds; stomach-turning acceleration; a cyclone of serving trays and carry-on luggage; fierce winds and extreme cold.

There was no surviving the final impact. All 298 would die.

As flaming wreckage fell just outside the town of Hrabove, it took under five minutes for the first reports to appear online. A local witness described the moment she would never forget: the bodies “just fell very, very hard to the ground.”

As the story rocketed across the internet, each side in the war zone that the plane had fallen into—the Ukrainian government and Russian-backed separatists—quickly blamed the other for the tragedy. Although social media exploded with theories, actual facts were hidden in a swirling fog. The rebels barred international investigators from visiting the crash site, preventing any independent examination for the next two weeks. It seemed that whoever had just killed 298 civilians would have ample time to disappear.

What they hadn’t counted on was a soft-spoken former World of Warcraft addict. Sitting 2,000 miles away at his computer, he already had access to all the evidence he would need.

Three years earlier, Eliot Higgins had been a stay-at-home dad, doting on his infant daughter in their cozy Leicester, England, home. Deciding he was spending too much time online playing video games and commenting on news stories, he turned to channel his interests into something more useful—starting a blog about the Syrian civil war that had just begun. He took the handle “Brown Moses,” from one of the iconoclastic rocker Frank Zappa’s more obscure songs, which asked, “What wickedness is dis?”

Yet Higgins had never even been to Syria. He didn’t speak Arabic. By his own admission, his knowledge of any type of conflict was limited to what he’d seen in Rambo movies. Indeed, he rarely left his house. He didn’t have to. Thanks to millions of social media accounts, the Syrian civil war came to him.

His gifts were patience and diligence. His weapons were YouTube and Google Maps. Higgins taught himself how to find and track weapon serial numbers; how to use landmarks and satellite imaging to trace someone’s steps; how to combine and catalog a stream of tens of thousands of videos. Soon enough, Higgins was charting each new development in a twisting, chaotic conflict. He uncovered hidden rebel weapon supply lines. He built a mountain of evidence to show that the Syrian dictator Bashar al-Assad had used nerve gas on his own people. From modest Blogspot beginnings, Brown Moses was soon rivaling not just the professional news media in his reporting, but some government intelligence agencies as well.

But this was only his first act. As interest in his unusual methods grew, Higgins launched a crowdfunding campaign for a new kind of online project dedicated to “citizen investigative journalists.” The organization was dubbed Bellingcat. The name was taken from the old fable in which a group of mice conspire to place a bell around a cat’s neck, so that they might always be warned of its approach.

Bellingcat had barely launched when flight MH17 fell from the sky. The first cat had come stalking.

One day after the tragedy, with international media awash in speculation and finger-pointing, Bellingcat published its first report. It was a straightforward summary of social media evidence to date, focused on sightings of a Russian Buk surface-to-air missile launcher that had been prowling near the crash site at the time of the tragedy. On July 22, Higgins posted a follow-up, superimposing images of wreckage against pictures of the intact plane, tracing the pattern of shrapnel damage. He drew no firm conclusions, pushing back against sketchy accounts posted by both Russians and Ukrainians. Bellingcat would only report the facts, as the digital bread crumbs revealed them.

It took a certain kind of person to excel in this work. Social media forensics requires endless focus and an attention to detail that sometimes borders on unhealthy. “I played a lot of role-player games,” Higgins explained. “Believe me, there are a lot of obsessive people out there who could probably put their passions to a more productive use.” A diverse crew of obsessive volunteers joined the hunt, an international online collective forming. They ranged from a Finnish military officer who knew Russian weapons, to Aric Toler, an American volunteer from North Carolina, who sheepishly told us how his main qualification was just being “good at the internet.” He’d spend hours each day tumbling through obscure Russian social media channels, surfacing only for occasional coffee breaks and visits to the Chipotle next door. His in-laws thought he was wasting his time online. They didn’t realize Toler was investigating war crimes.

The Bellingcat team soon tracked down multiple images and videos that showed a Buk missile launcher in the vicinity of MH17’s flight path on the day of the tragedy, clearly within separatist territory. But the team noticed something even more telling. In photographs posted from before the time of the plane crash, the vehicle carried four missiles; in photographs taken soon after the crash, only three. They’d found their smoking gun.

But then the trail seemed to go cold. Although they could find plenty of other online photos of Russian-manned Buks deployed to aid the rebels, they couldn’t find any additional matches for this particular vehicle. The images also showed that a shell game was being played. The painted vehicle number had been changed both before and after the July 17 event, and it could easily be changed again.

The breakthrough came when the analysts shifted their gaze lower. They realized that every Buk vehicle had a rubber skirt to help stop its treads from throwing up mud and dirt. Because each vehicle has its own particular driving history, each rubber skirt has a unique pattern of wear and tear. Now the Bellingcat team had the equivalent of a fingerprint to hunt for across every photograph and snippet of video that came out of eastern Ukraine.

They soon located photos of the Buk that had shot the missile in a convoy filmed crossing from Russia into Ukraine on June 23—and leaving again on July 20. They were then able to trace backward, finding the unit to which the Buk in the convoy belonged: the 2nd Battalion of the 53rd Anti-Aircraft Missile Brigade of the Russian army. Bellingcat posted their findings online, mapping out the odyssey of the weapon that had killed 298 people, as well as showing its Russian origin.

The weapon and even its unit had been found, but who had pulled the trigger? Here the answer was provided by the shooters themselves. Searching through Russian soldiers’ profiles on VKontakte, or VK (essentially a Russian version of Facebook), the Bellingcat team found images of military equipment, dour group photographs, and hundreds of angsty selfies. One conscript had even snapped a picture of an attendance sheet for a 2nd Battalion drill shortly before its deployment to Ukraine.

It wasn’t just the soldiers who opened the lens to their war, but also their friends and families. The Bellingcat team found particular value in an online forum frequented by Russian soldiers’ wives and mothers. Worried about their loved ones, they traded gossip about the deployments of specific units, which also proved an intelligence gold mine.

After nearly two years of research, Bellingcat presented its findings to the Dutch tribunal that had been tasked with bringing the killers to justice. It included the names, photographs, and contact information of the twenty soldiers who the data showed had been manning the missile system that had shot down flight MH17. It was an extraordinary feat, accomplished using only what was available on the internet. It was also damning evidence of Russian participation in a war crime.

Through their stubborn and breathtakingly focused investigation of the MH17 case, Eliot Higgins and Bellingcat displayed the remarkable new power of what is known as “open-source intelligence” (OSINT).

With today’s OSINT, anyone can gather and process intelligence in a way that would have been difficult or impossible for even the CIA or KGB a generation ago. One OSINT analyst explained to us just how simple it can be, through a story of his pursuit of Iranian arms smuggling. He began by searching for some common weapons-related words in Farsi, courtesy of Google Translate. Soon he discovered an article that profiled a young Iranian CEO, who had launched a company specializing in drones. The journalist also found a video online that showed the CEO’s face. This allowed him to locate the CEO in a registry of Iranian aerospace professionals, which, in turn, yielded an email address as well as phone and fax numbers. Translating the CEO’s name into Farsi and searching Facebook (cross-referencing it with the email address), he quickly found the CEO’s account. It was confirmed with a perfect facial match. He was then able to trace the CEO’s movements, including a trip to Malaysia, where there is a market for drone parts. He also found that, despite working for a theocratic state, the Iranian CEO had a notable friendship with a charming young woman there, who liked to wear bunny ears and a six-inch skirt. Her own posts, in turn, revealed her to be an alumna of the “International Sex School.”

He ended the search there. In just one hour of digging, he’d been able to compile a list of leads that might once have taken an intelligence agency months to find, as well as to tease out a fertile opportunity for blackmail. Although he’d begun his work as a reporter, in the end it was tough to shake the feeling that he’d become something else entirely: a spy.

In some cases, these new OSINT analysts don’t even have to be human. GVA Dictator Alert is an algorithm whose sole purpose is tracking the flights of dictators into and out of Geneva, Switzerland—a favorite destination for money laundering and other shady business dealings. The program does its job unerringly, scanning real-time aviation data for aircraft registered to repressive governments. Whenever an autocrat arrives to check on their money, GVA Dictator Alert instantly squawks the information across Twitter. As the bot’s creator explained, “It’s a cool idea to know that every time the front wheel of their plane touches the tarmac in Geneva, there’s a tweet saying ‘hi, you are here, and now it’s public.’”

In revealing these secrets, OSINT showed how its ability to unveil what was once hidden can be a powerful potential force for good. It can not only catch people cutting corners (quite literally, OSINT analysts found that one of every five racers in the 2017 Mexico City Marathon cheated, including several politicians hoping to tout their endurance), but also shine a light on the world’s worst crimes. The manner in which Hitler and Pol Pot were able to kill en masse, unbeknownst to the wider world, is simply not possible today. The Bellingcat team, for instance, would use the very same approach it had brought to detecting war crimes in Ukraine to documenting the use of chemical weapons in Syria. Indeed, both technology and international law crossed a new frontier in 2017, when the International Criminal Court indicted Mahmoud Al-Werfalli for a series of mass killings in Libya. He was the first person ever to be charged with war crimes based solely on evidence found on social media.

Yet the very same techniques can also aid evil causes. Terrorists can reconnoiter potential targets without ever visiting them in person; they can tap into a global network of bomb makers and weapons experts without leaving their homes. The OSINT revolution also enables wholly new categories of crime. In Colombia, Mexico, and Venezuela, targets for kidnapping are selected through information gathered from their social media accounts. In some “virtual kidnappings,” the actual kidnapping is skipped entirely as the criminals extort ransom from friends and family, knowing when the “victim” is out of contact and unable to respond.

OSINT can even offer a glimpse into previously impenetrable worlds. After decades of dark secrecy, the modern “Lords of War”—arms dealers—have embraced social media along with everybody else. Scouring Libyan Facebook groups, you can track hundreds of arms trades each month, as dealers advertise and negotiate shipments of everything from anti-aircraft missiles to heavy machine guns. By monitoring these sales, you can spot not just looming battles (evidenced by who is stocking up) but also evidence of policy failure. Many of the thousands of weapons traded online by Middle Eastern Facebook users, for instance, can be traced to stocks of weapons given to the Iraqi army and Syrian rebels by the CIA and U.S. military.

At its most promising, the OSINT revolution doesn’t just help people parse secrets from publicly accessible information; it may also help them predict the future. Predata is a small company founded by James Shinn, a former CIA agent. Shinn modeled his unique service on sabermetrics, the statistics-driven baseball analysis method popularized by Michael Lewis’s book Moneyball. “By carefully gathering lots and lots of statistics on their past performance from all corners of the internet, we are predicting how a large number of players on a team will bat or pitch in the future,” Shinn explains. In this case, however, the statistics his firm mines are tens of millions of social media feeds around the world. But instead of predicting hits and strikeouts, it predicts events like riots and wars.

Predata uses such mass monitoring to discern online patterns that might be used to project real-world occurrences. Each Sunday, it sends out a “Week Ahead” mailer, breaking down the statistical likelihood of particular contingencies based on web monitoring. Billion-dollar Wall Street hedge funds are interested in any hints of unrest that might move markets. U.S. intelligence agencies are interested in signs of looming terror attacks or geopolitical shifts. For instance, North Korean missile and nuclear bomb tests have been predicted by analysts studying the correlation between past tests and social media chatter, using meta-analysis of online conversations and website visits. The world of social media is becoming so revelatory that it can even help someone to anticipate what will happen next.

TRUE BELIEVER

“The exponential explosion of publicly available information is changing the global intelligence system . . . It’s changing how we tool, how we organize, how we institutionalize—everything we do.”

This is how a former director of the U.S. Defense Intelligence Agency (DIA) explained to us how the people who had once owned and collected secrets—professional spies—were adjusting to this world without secrets.

OSINT has a long history, being first separated from the classic spy craft of coaxing and interrogation (known as human intelligence, or HUMINT) and the intercept of confidential communications (signals intelligence, or SIGINT) during World War II. The breakthrough came when Allied analysts with the Office of Strategic Services (the predecessor of today’s CIA) discovered that they could figure out the number of Nazi casualties by reading the obituary sections of German newspapers that were available in neutral Switzerland. By war’s end, these analysts were cataloging roughly 45,000 pages of European periodicals each week. America also launched the Foreign Broadcast Monitoring Service (renamed the Foreign Broadcast Information Service, or FBIS), which transcribed 500,000 words of radio broadcasting each day.

Through much of the Cold War, U.S. intelligence agencies collected OSINT on a massive scale. The U.S. embassy in Moscow maintained subscriptions to over a thousand Soviet journals and magazines, while the FBIS stretched across 19 regional bureaus, monitoring more than 3,500 publications in 55 languages, as well as nearly a thousand hours of television each week. But intelligence chiefs traditionally put little faith in what this mass of free data yielded, rarely giving it the same weight as the other sources of information. Part of their skepticism arose because the information was so readily available (if it could be acquired so easily, how could it be valuable?), and part of it was because they suspected trickery (anything willingly shared by the Soviet Union must be a lie).

Ultimately, the FBIS was undone by the sheer volume of OSINT the internet produced. In 1993, the FBIS was creating 17,000 reports a month; by 2004, that number had risen to 50,000. In 2005, the FBIS was shuttered. Information was simply spreading too quickly online, in too many forms, for it to keep up. There was also no reason for the U.S. government to work so hard. For years, analysts had labored to maintain a sprawling, updated encyclopedia on the regions of the Soviet Union. Now there was Wikipedia.

However, a few forward-thinking intelligence officers dared to take the next big cognitive leap. What if OSINT wasn’t losing its value, they asked, but was instead becoming the new coin of the realm? The question was painful because it required setting aside decades of training and established thinking. It meant envisioning a future in which the most valued secrets wouldn’t come from cracking intricate codes or the whispers of human spies behind enemy lines—the sort of information that only the government could gather. Instead, they would be mined from a vast web of open-source data, to which everyone else in the world had access. If this was true, it meant changing nearly every aspect of every intelligence agency, from shifting budget priorities and programs to altering the very way one looked at the world. But the intelligence expert we interviewed felt it was a crucial change that had to be made.

“Publicly available information is now probably the greatest means of intelligence that we could bring to bear,” he told us. “Whether you’re a CEO, a commander in chief, or a military commander, if you don’t have a social media component . . . you’re going to fail.”

The expert we consulted was Michael Thomas Flynn.

Flynn joined the U.S. military in 1981, at the height of the Cold War. He built his career in Army intelligence, rising through the ranks. After 9/11, he was made director of intelligence for the task force that deployed to Afghanistan. He then assumed the same role for the Joint Special Operations Command (JSOC), the secretive organization of elite units like the bin Laden–killing Navy SEAL team that had been outed by social media in Abbottabad. In this position, trying to track down the terror cells of Al Qaeda in Iraq (AQI), Flynn realized that his operatives had to look elsewhere, out in the open, for clues to where the enemy was hiding. And they had to do it faster than ever before.

As he explained, U.S. special operations forces were commandos without equal, “the best spear fishermen in the world.” But in order to beat an adversary that recruited so rapidly and blended so easily with the civilian population, the commandos would have to become “net fishermen.” They would eschew individual nodes and focus instead on taking down the entire network, hitting it before it could react and reconstitute itself. As Flynn’s methods evolved, JSOC got better, capturing or killing dozens of terrorists in a single operation, gathering up the intelligence, and then blasting off to hit another target before the night was done. Eventually, the shattered remnants of AQI would flee Iraq for Syria, where they would ironically later reorganize themselves into the core of ISIS.

Flynn’s career took off. He was promoted to three-star general and, in 2012, appointed to lead the DIA, the agency charged with centralizing intelligence across the entire U.S. military. Although he had no experience commanding such a large organization (the DIA numbered some 17,000 employees), Flynn was eager to translate his ideas into action. He envisioned not just reform of the DIA, but a wider reorganization of how the entire intelligence system worked in the twenty-first century. It was time to shift what was being collected and how. Before the rise of social media, he explained, 90 percent of useful intelligence had come from secret sources. Now it was the exact opposite, with 90 percent coming from open sources that anyone could tap.

Flynn sought to steer the agency in a new direction, boosting OSINT capabilities and prioritizing the hiring of computational analysts, who could put the data gushing from the digital fire hose to good use. He expected an uphill battle. OSINT, he explained, had only recently stopped being the “unwanted pregnancy” of military intelligence. Now, at best, it was a “redheaded stepchild.”

He didn’t realize the shake-up would prove a bridge too far. Flynn’s aggressive moves alarmed the DIA’s bureaucracy—not least because it threatened their own jobs. The agency was soon mired in chaos. Flynn’s leadership was also questioned, his grand vision undermined by poor management. Just a year and a half after his term began, Flynn was informed that he was being replaced. He was forced into retirement, leaving the Army after thirty-three years of service.

If that were the whole story, Flynn’s legacy might be one of a forward-thinking prophet of the social media revolution who paid the price for seeking change. But his tale didn’t end there.

Flynn didn’t take his dismissal well. After he left the military, he channeled his energy into media appearances and speaking engagements, as well as building a consulting business. He was originally marketed as the general who had seen the future. But he quickly became better known as a critic of the Obama administration that had fired him, angrily denouncing it for betraying him and the nation. This brought celebrity and money well beyond what Flynn had made in military service, but also new entanglements. His firm signed a shady $530,000 deal with a company linked to the Turkish government, which became doubly questionable when Flynn failed to register as a foreign lobbying agent. He accepted $45,000 to speak at a glitzy Russian government–sponsored gala in Moscow. Photos of the ex-general sitting next to Vladimir Putin at the dinner shocked many in the U.S. security establishment.

Most important, Flynn’s rising celebrity came to the attention of Donald Trump, who had just announced his run for office. Their first meeting was supposed to last thirty minutes. When it ended ninety minutes later, the former intelligence officer walked away with new insight into the future. “I knew he was going to be President of the United States.”

Now the angry general became Trump’s fiercest campaign surrogate, bestowing the inexperienced candidate with much-needed national security credibility. He used his old Army rank as a weapon, relentlessly attacking Trump’s rivals. In doing so, however, Flynn began to dive deeply into the online world that he’d previously just observed. The result wasn’t pretty.

Flynn had started his personal Twitter account, @GenFlynn, in 2011 with a tweet linking to a news article on Middle East politics. Not a single person had replied or retweeted it. But as he entered politics, Flynn’s persona changed dramatically. His feed pushed out messages of hate (“Fear of Muslims is RATIONAL,” he fumed in one widely shared tweet), anti-Semitism (“Not anymore, Jews. Not anymore,” referring to the news media), and one wild conspiracy theory after another. His postings alleged that Obama wasn’t just a secret Muslim, but a “jihadi” who “laundered” money for terrorists, that Hillary Clinton was involved with “Sex Crimes w Children,” and that if she won the election, she would help erect a one-world government to outlaw Christianity. To wild acclaim from his new Twitter fans, Flynn even posted on #spiritcooking, an online conspiracy theory that claimed Washington, DC, elites regularly gathered at secret dinners to drink human blood and semen. That message got @GenFlynn over 2,800 “likes.”

It was a remarkable turn for the once-respected intelligence officer. Just a few months earlier, he had cautioned us about the internet, “Now it’s a matter of making sure that the accuracy matters . . . You combine your judgment, your experience, your analysis along with the valuable data you get.”

Despite the online madness that violated his advice (or perhaps because of it), things seemed to work out well for the general. When Trump won the election, Flynn was named to the position of national security advisor, one of the most powerful jobs in the world. His first tweet in the new role proclaimed, “We are going to win and win and win at everything we do.”

The winning didn’t last long.

Within a few weeks, Flynn would be fired, done in by a web of mistruths regarding his contact with Russian government officials. He was the shortest-serving national security advisor in American history. Within the year, in a plea bargain with the U.S. Department of Justice, Flynn would admit to making, in his words, “false, fictitious, and fraudulent statements.”

As it all played out, we were reminded of one more piece of wisdom Flynn had imparted to us before his downfall. He’d spoken of the importance of piercing through the “fog” of the modern information environment; of getting to the “golden nuggets” of actionable intelligence that lurked in the mists of social media. The right bit of data was already out there, he explained. You just had to know where to look.

The general was right. The internet has indeed exposed the golden nuggets—the truth—for anyone to find. But, as his story also shows, scattered among these bits of truth is pyrite, “fool’s gold,” cleverly engineered to distract or even destroy us. It is harder than ever to keep a secret. But it is also harder than ever to separate the truth from lies. In the right hands, those lies can become powerful weapons.