The Perfect Weapon

by David E. Sanger

  In mid-January of 2011, we felt we had enough information to publish our first story about who had been behind the Stuxnet attacks. In a Sunday article, we laid out the compelling evidence that the United States and Israel had produced the malware together in order to slow Iran’s nuclear progress. The story was full of details and markers that took the code right to the gate of Fort Meade, where the NSA is located, but upon publication there was no political outcry, no investigation. That would come more than a year later.*2

  But even after we published our account it was clear there were major questions left unanswered: Had this been a small operation gone awry, or a large, well-hidden effort? Assuming the United States and Israel had combined forces to design this enormously complex cyberweapon, who had given the go-ahead? After all, we knew that in the United States only the president could authorize offensive cyber action, just as he had to provide the launch codes for nuclear weapons.

  And if Olympic Games was a sign of where American covert action was headed, were we ready as a nation to open this Pandora’s box? Once opened, could it ever be closed again?

  * * *

  —

  The discovery that Israel had built a replica of the Natanz plant drove home how central a role the Israelis had played in developing the Stuxnet malware. The more sources I interviewed, the more it became clear that the cyber program widened a divide between Prime Minister Benjamin Netanyahu and his brilliant, short, bald spymaster, Meir Dagan. In Dagan’s younger days in the Israeli military, he had led squads that hunted down Palestinian militants. Ariel Sharon, the Israeli prime minister who had been Dagan’s commander and mentor, famously if crudely declared that “Dagan’s specialty is separating an Arab from his head.” It was a brutal description, even in the macho world of the Mossad, Israel’s best-known intelligence agency, which Dagan ultimately led for nine years—an extraordinarily long tenure. While Dagan pretended to dismiss the stories as mythmaking, he nonetheless seemed to revel in them.

  But the mythmaking ignored the fact that it wasn’t only Arabs whom Dagan had in his sights. Many observers suspected Dagan’s hand in the killing of Iranian nuclear scientists, who were assassinated while driving to work in Tehran traffic after motorcyclists pulled up and attached “sticky bombs” to their car doors before speeding off. If Dagan were indeed behind the killings, it would be in keeping with his view that an Iran armed with a nuclear weapon was truly an existential threat to Israel. Indeed, to talk to Dagan for five minutes was to discover a man who viewed the world through the lens of the Holocaust. On his desk, he kept a photograph of his grandfather kneeling on the ground before his Nazi captors moments before he was killed. It was Dagan’s personal “never again” memento that seemed to explain the determination with which he organized the elimination of Israel’s enemies.

  Dagan made no secret that he never hesitated to send Mossad agents out to kill. Yet when one of those missions went bad—his agents were caught on tape entering and leaving a hotel in Dubai just before and after the 2010 killing of a senior official of Hamas, the Islamist Palestinian group—it was the beginning of the end of his career. The images of the Israeli agents, dressed casually in tennis gear as they entered and left the hotel, played again and again on television. But as his time as the chief of the Mossad dwindled down, Dagan wanted to be remembered instead for managing an operation that was, in his mind, a complete success: the malware attack that crippled Natanz.

  Despite Dagan’s public reputation as a brutal spymaster who had killed many Arabs in his younger days and ordered the deaths of many more from Mossad headquarters, he was far more strategically savvy than most Israelis knew. Internally, he was increasingly vocal that bombing Iran was madness—it would simply drive the nuclear program further underground. That program would then come back, bigger and more advanced than before. Dagan devoted his last years in office to dissuading Prime Minister Netanyahu from an air attack. “The use of [military] violence would have intolerable consequences,” Dagan later told Israeli investigative journalist Ronen Bergman. “If Israel were to attack, Khamenei would thank Allah,” he said, referring to the Iranian Supreme Leader. “It would unite the Iranian people behind the project and enable Khamenei to say that he must get himself an atom bomb to defend Iran against Israeli aggression.”

  All of which meant that by 2010 Dagan was under tremendous pressure to show Netanyahu that a more covert, more sophisticated approach to crippling the Iranian program could succeed.

  Dagan and I never met when he was in office. But I was determined to change that after I heard that at one of his retirement parties many of the toasts and jokes made oblique references to the cyberattacks on Natanz. The insiders got the drift; others were left to wonder what everyone else was laughing about.

  We first talked in 2011 a few months after Dagan had been ousted from his job by Prime Minister Netanyahu. It was clear he was still bitter about his ouster. He variously derided Netanyahu as a terrible manager and an incompetent warrior. Rightly or wrongly, Dagan believed that Netanyahu had gotten rid of him because the Mossad chief, like other Israeli intelligence leaders, opposed efforts by the prime minister to bomb Iran’s nuclear facilities.

  “Bombing would be the stupidest thing we could do,” Dagan told me. This was not like striking Iraq’s Osirak nuclear reactor in 1981 or Syria’s reactor in 2007. He believed Iran’s program was simply too sprawling; they were not about to repeat their neighbors’ mistakes. So while an air attack on Iran’s facilities “might make me feel good,” Dagan said to me one afternoon, it would provide an illusory solution. The satellite photographs, he said, would show Iran’s facilities flattened, and everyone would cheer. But within months, he predicted, those facilities would be rebuilt so deep as to be impermeable to a second strike. And that, he thought, would be disastrous for the state of Israel.

  It was fine to try to slow Iran’s progress, said Dagan. But if Israel attempted to destroy the country’s nuclear facilities in an overt attack, it would ensure a nuclear Iran. There had to be a better way.

  A cyberweapon, in his view, was the way out of the conundrum. In our early meetings, Dagan was coy about his role in the development of Stuxnet, even when I mentioned that I had heard about his participation in secure videoconference debates about next steps in the attacks. He didn’t know much about computers, he often replied to me with a smile, as if that exonerated him from the role we both knew he played.

  Yet over time, as he grew sicker from a failing liver transplant, he edged closer to describing what happened, and why. In our handful of conversations over the years, he sprinkled phrases like “if we did it” into many sentences so that he could explain his underlying logic without violating his oath to maintain the secrecy around the Mossad’s covert operations. He talked about how Israel’s technology made it enormously difficult for the Iranians to figure out the origin of the attack. The operation against Iran was a model of how Israel should defend itself in the future, he said. Gone were the days of open demonstrations of military might that invited retaliation, escalation, and international condemnation. Gone were the days of occupying territory. The defense of Israel, he insisted, required subtlety and indirection.

  “I doubt he knew the first thing about how you write a string of code,” one American who dealt with Dagan often told me. “But he knew a lot about how you play with an enemy’s mind.” And he was convinced that it was the intelligence services that would end Iran’s nuclear program, not the air force. It was a mind-set that put Dagan and Cartwright in the same place. And many of Dagan’s fellow intelligence chiefs, once they left office, claimed they backed his arguments.

  I never heard Dagan directly admit his role in the cyberattacks. But he hinted they had been designed as much to divert Netanyahu from stumbling into a Mideast war as to stop the Iranians from enriching uranium. “I don’t trust him,” Dagan said of Netanyahu. It was in Netanyahu’s interest, he told me,
to portray the Iranians as irrational zealots who would use their bomb against Israel. But Dagan looked at the Iranians and saw a group of mullahs mostly interested in staying in power, rather than starting a suicidal war.

  There was a reason, Dagan told me, that Bush would not give the biggest bunker-busting bombs to Netanyahu. “He was afraid Bibi would use them,” Dagan told me. “And so was I.”

  That fear explained Dagan’s enthusiasm for using cyberweapons against Tehran. It was a way to set the nuclear program back. Perhaps more important, Bush and Obama were able to argue to Netanyahu that there was no reason to bomb while the cyberattacks were working.

  The last time I saw Dagan, he chewed me out for what I had written about Olympic Games. But unlike his American counterparts, he complained that I had written too little, not too much.

  “You missed a major part of the story,” he said, arguing that the Americans had received far too much credit, and the Israelis—and by extension Dagan himself—had received not nearly enough. I had been seduced by Americans who were intoxicated with advertising their own success, he insisted one evening, rather than giving credit to an ally—he carefully didn’t say which one—that had done the heavy lifting, gotten the code into the centrifuges, and revised it as necessary.

  I would be happy to tell more of Israel’s side of the story, I told him, but he’d have to be more explicit about the operational details to prove his point. He smiled—a smile of disgust.

  “I am an old man,” he said, “and I am sick. I don’t want to spend my last days in jail.”

  * * *

  —

  By the end of 2011, after dozens of interviews, I had pieced together the highlights of the story of the strategy and debates swirling around the decision to unleash Stuxnet—or at least as many of them as I could gather, given the layers upon layers of secrecy involved. After consulting editors, and the New York Times in-house counsel, it was time to go to the Obama White House to see if they were ready to talk—both about what had happened and about any national security concerns they might have about publishing the details. As in all such cases, I made clear that the Times, and the Times alone, would decide what to publish. But if there were risks to ongoing operations or lives, we needed to discuss them now, not after publication.

  My first visit was to Benjamin Rhodes, the former novelist and graceful speechwriter who eventually handled a portfolio of diplomatic issues for Obama, including the opening of Cuba. It was his job to deal with reporters who came to the administration with complex, sensitive national security stories and to decide how, if at all, the White House would respond. Without getting into the details of what happened, he suggested that I go see General Cartwright. It made sense, since Cartwright’s term in office spanned the Bush and Obama administrations, and he had been at the center of all offensive cyber debates and understood their sensitivities. Cartwright had retired from the marines after he was passed over for chairman of the Joint Chiefs in 2011 and knew the history of the US military’s development of a cyber arsenal better than anyone else.

  I knew Cartwright from his days on the Joint Chiefs and had attended conferences where he had discussed the strategic challenges of the new age of cyber conflict. In my reporting on Olympic Games, his name came up often as the man who had tutored Obama about how the Stuxnet worm worked (though it had not yet been given the name “Stuxnet”) and rolled out the “horse blanket” diagram of Natanz to bring Obama up to date.

  But Cartwright’s direct line to Obama had grated on Robert Gates, the secretary of defense, and Mike Mullen, the chairman of the Joint Chiefs. On a variety of issues they believed he manipulated the Pentagon system, or went around the chain of command. It didn’t help that Cartwright had not spent time in Iraq and Afghanistan. When Mullen was ready to retire, the two successfully argued against promoting Cartwright into Mullen’s role. Suddenly, the man who was among the first to sketch out how the United States could create a dedicated military command to deal with a new dimension of warfare was cast out. With him, I discovered later, went some of the most creative strategic thinking about the use of cyber in offense and defense.

  Since retiring, Cartwright had taken a chair at the Center for Strategic and International Studies and signed on with a handful of defense firms, including Raytheon, the maker of missile defenses and defense electronics. Carefully, he had begun to speak out against the secrecy surrounding America’s new cyber arsenal, arguing that if the United States wanted to create cyber deterrence it was going to have to show a bit of its capability. “You can’t have something that’s a secret be a deterrent,” I heard him say in more than a few public forums. “Because if you don’t know it’s there, it doesn’t scare you.”

  He was right, and with the utmost care, the Pentagon began to slip a few lines into public testimony acknowledging that it possessed offensive cyber capabilities. It was a little like acknowledging that the sun rose in the morning. Still, it wasn’t an enormously welcome message in the intelligence world, which feared a slippery slope toward divulging what those weapons were and how they were used. Meanwhile, Cartwright was also making a case that the United States could survive quite nicely, thank you, with far fewer nuclear weapons—an argument that carried extra weight coming from the former head of Strategic Command. Again, he was right. But his argument didn’t exactly win him friends among his old Pentagon colleagues, who rarely met a weapons system they didn’t like.

  I took Rhodes’s advice and called Cartwright. By the time I went to see him, I not only knew the outlines of the story about Olympic Games, I had already drafted them into two chapters of a book about Obama’s first term, describing Olympic Games in as much detail as we could dig up at the time. The book was due to be released in months, and the manuscript was already being edited. That fact became significant later on, when the FBI—whose special agents must have never encountered a book-production schedule before—wrongly concluded that Cartwright was the source of the tale.

  My goal in seeing Cartwright was twofold: to check that I had the history and implications right, and to get an independent view of whether any details I was reporting could jeopardize American national security. Cartwright knew that I had been sent by the White House, and saw himself as part of the effort to dissuade me from publishing any details of the operation that might aid an American adversary. He made it clear that he could not discuss classified details. Yet as it turned out when the FBI went looking for a “leaker”—as if there were a single one—we were both a little naïve. For doing what he thought was right, he later paid an awful price, for which I feel enormously guilty to this day.

  * * *

  —

  It was just a few days later that I made my trek out to the CIA headquarters to visit Michael Morell in his seventh-floor office—surrounded by three decades of memorabilia from his career, including artifacts from the raid that killed Osama bin Laden. Morell was close to President Obama, and I knew that if the administration was going to push back against the story, he would be the man to do it.

  We began pacing through my reporting and the story I was preparing to tell. I ticked through the forensics that led experts to identify the United States and Israel, the carcasses of blown-apart centrifuges found by international inspectors, the mock-ups of Natanz built by the Israelis at Dimona and by the United States in Tennessee. I described the Situation Room debates in which Morell participated. He cautioned against a few assertions and argued with a few conclusions. At a few moments in the story he slowed down, taking notes and suggesting that he might ask that I remove references to certain techniques that the agency used to get malware into target computers and networks. (Curiously, a few weeks later, he asked that a reference to one of those techniques be restored. Though he offered no explanation, clearly the agency had moved on to other methods and wanted to keep the Iranians thinking that the old techniques were still in use.)

  In the en
d, Morell asked for only a handful of deletions, mostly technical details that focused on how the United States put “beacons” and malware into foreign systems and networks. None was essential to telling the story of the most sophisticated state-sponsored cyberattack in history. “You agreed to just about everything we asked for,” he acknowledged later on, even while still objecting to the fact that we were publishing anything at all about an American covert operation.

  But none of that mattered when the story was published. Republicans who were trying to cast Obama as weak on terrorism—not easy after the killing of bin Laden—accused the White House of leaking the story, along with an unrelated story in the Times about the president’s role in approving a “kill list” of terrorists to be targeted by drones.

  “We know the leaks have to come from the administration. And so we’re at the point where perhaps we need an investigation,” said Sen. John McCain. He called the story part of “a pattern in order to hype the national security credentials of the president and every administration does it. But I think this administration has taken it to a new level.”

  Obama himself performed a delicate dance: He couldn’t confirm the story, of course, or deny it, but he wanted the world to know he wasn’t the source. “I’m not going to comment on the details of what are supposed to be classified items,” he said with a hardness in his voice a few days after the details about the White House origins of Olympic Games were published. “When this information, or reports, whether true or false, surface on the front page of newspapers, that makes the job of folks on the front lines tougher and it makes my job tougher—which is why, since I’ve been in office, my attitude has been zero tolerance for these kinds of leaks and speculation.