A year into Donald J. Trump’s presidency, his defense secretary, Jim Mattis, sent the new commander-in-chief a startling recommendation: with nations around the world threatening to use cyberweapons to bring down America’s power grids, cell-phone networks, and water supplies, Trump should declare he was ready to take extraordinary steps to protect the country. If any nation hit America’s critical infrastructure with a devastating strike, even a non-nuclear one, it should be forewarned that the United States might reach for a nuclear weapon in response.
Like most things in Washington, the recommendation leaked immediately. Many declared it a crazy idea, and wild overkill. While nations had turned their cyberweapons against each other dozens of times in recent years, no attack had yet been proven to cost a human life, at least directly. Not the American attacks on Iran’s and North Korea’s weapons programs; not the North Korean attacks on American banks, a famed Hollywood studio, and the British healthcare system; not the Russian attacks on Ukraine, Europe, and then the core of American democracy. That streak of luck was certain to end soon. But why would Donald Trump, or any of his successors, take the huge risk of escalating a cyberwar by going nuclear?
The Pentagon’s recommendation, it turned out, was the prelude to other proposals—delivered to a president who values toughness and “America First”—to use the nation’s powerful cyberweapons far more aggressively. But it was also a reminder of how quickly the fear of devastating cyberattacks has moved from the stuff of science fiction and Die Hard movies to the center of American defense strategy. Just over a decade before, in 2007, cyberattacks were missing entirely from the global “Threat Assessment” that intelligence agencies prepare each year for Congress. Terrorism topped that list, along with other post-9/11 concerns. Now that hierarchy has been reversed: For several years a variety of cyber threats, ranging from a paralyzing strike on the nation’s cities to a sophisticated effort to undercut public confidence in its institutions, has appeared as the number one threat on the list. Not since the Soviets tested the Bomb in 1949 had the perception of threats facing the nation been revised so quickly. Yet Mattis, who had risen to four-star status in a career focused on the Middle East, feared that the two decades spent chasing al Qaeda and ISIS around the globe had distracted America from its most potent challenges.
“Great power competition—not terrorism—is now the primary focus of US national security,” he said in early 2018. America’s “competitive edge has eroded in every domain of warfare,” including the newest one, “cyberspace.” The nuclear strategy he handed Trump gave voice to an inchoate fear among many in the Pentagon that cyberattacks posed a threat unlike any other, and one we had completely failed to deter.
The irony is that the United States remains the world’s stealthiest, most skillful cyberpower, as the Iranians discovered when their centrifuges spun out of control and the North Koreans suspected as their missiles fell out of the sky. But the gap is closing. Cyberweapons are so cheap to develop and so easy to hide that they have proven irresistible. And American officials are discovering that in a world in which almost everything is connected—phones, cars, electrical grids, and satellites—everything can be disrupted, if not destroyed. For seventy years, the thinking inside the Pentagon was that only nations with nuclear weapons could threaten America’s existence. Now that assumption is in doubt.
In almost every classified Pentagon scenario for how a future confrontation with Russia and China, even Iran and North Korea, might play out, the adversary’s first strike against the United States would include a cyber barrage aimed at civilians. It would fry power grids, stop trains, silence cell phones, and overwhelm the Internet. In the worst-case scenarios, food and water would begin to run out; hospitals would turn people away. Separated from their electronics, and thus their connections, Americans would panic, or turn against one another.
The Pentagon is now planning for this scenario because it knows many of its own war plans open with similarly paralyzing cyberattacks against our adversaries, reflecting new strategies to try to win wars before a shot is fired. Glimpses of what this would look like have leaked out in recent years, partly thanks to Edward J. Snowden, partly because a mysterious group called the Shadow Brokers—suspected of close links to Russian intelligence—obtained terabytes of data containing many of the “tools” that the National Security Agency used to breach foreign computer networks. It didn’t take long for some of those stolen cyberweapons to be shot back at America and its allies, in attacks whose bizarre-sounding names, like WannaCry, suddenly appeared in the headlines every week.
Yet the secrecy surrounding these programs obscures most public debate about the wisdom of using them, or the risks inherent in losing control of them. The government’s silence about America’s new arsenal, and its implications, poses a sharp contrast to the first decades of the nuclear era. The horrific scenes of destruction at Hiroshima and Nagasaki not only seared the national psyche, but they made America’s destructive capabilities—and soon Russia’s and China’s—obvious and undeniable. Yet even while the government kept the details classified—how to build atomic weapons, where they are stored, and who has the authority to order their launch—America engaged in a decades-long political debate about when to threaten to use the Bomb and whether to ban it. Those arguments ended up in a very different place from where they began: in the 1950s the United States talked casually about dropping atomic weapons to end the Korean War; by the eighties there was a national consensus that the US would reach for nuclear weapons only if our national survival was at stake.
So far, there has been no equivalent debate about using cyberweapons, even as their destructive power becomes more evident each year. The weapons remain invisible, the attacks deniable, the results uncertain. Naturally secretive, intelligence officials and their military counterparts refuse to discuss the scope of America’s cyber capabilities for fear of diminishing whatever narrow advantage the country retains over its adversaries.
The result is that the United States makes use of this incredibly powerful new weapon largely in secret, on a case-by-case basis, before we fully understand its consequences. Acts that the United States calls “cyber network exploitations” when conducted by American forces are often called “cyberattacks” when American citizens are the target. That word has come to encompass everything from disabling the grid, to manipulating an election, to worrying about that letter arriving in the mail warning that someone—maybe criminals, maybe the Chinese—just grabbed our credit cards, Social Security numbers, and medical histories, for the second or third time.
During the Cold War, national leaders understood that nuclear weapons had fundamentally changed the dynamics of national security, even if they disagreed on how to respond to the threat. Yet in the age of digital conflict, few have a handle on how this new revolution is reshaping global power. During his raucous 2016 presidential campaign, Trump told me in an interview that America was “so obsolete in cyber,” ignoring, if he was aware of it, that the United States and Israel had deployed the most sophisticated cyberweapon in history against Iran. More concerning was the fact that he showed little understanding of the dynamics of the grinding, daily cyber conflict now under way—the short-of-war attacks that have become the new normal. His refusal to acknowledge Russia’s pernicious role in the 2016 election, for fear it would undercut his political legitimacy, only exacerbates the problem of formulating a national strategy. But the problem goes far beyond the Trump White House. After a decade of hearings in Congress, there is still little agreement on whether and when cyberstrikes constitute an act of war, an act of terrorism, mere espionage, or cyber-enabled vandalism. Technological change wildly outpaces the ability of politicians—and the citizens who have become the collateral damage in the daily combat of cyberspace—to understand what was happening, much less to devise a national response. Making matters worse, when Russia used social media to increase America’s polarization in the 2016 election, the animus between tech companies and the US government—ignited by Snowden’s disclosures four years earlier—only deepened. Silicon Valley and Washington are now the equivalent of a divorced couple living on opposite coasts, exchanging snippy text messages.
Trump accepted Mattis’s nuclear recommendation without a moment of debate. Meanwhile the Pentagon, sensing Trump’s willingness to demonstrate overwhelming American force in cyberspace as in other military arenas, published a new strategy, envisioning an era of constant, low-level cyber conflict in which America’s newly minted cyber warriors would go deep behind enemy lines every day, attacking foreign computer servers before threats to the United States could materialize. The idea was classic preemption, updated for the cyber age, to “stop attacks before they penetrate our cyber defenses or impair our military forces.” Other proposals suggested the president should no longer have to approve every cyber strike—any more than he would have to approve every drone strike.
In the chaos of the Trump White House, it was unclear how these weapons would be used, or under what rules. But suddenly we are in new territory.
Cyber conflict remains in the gray area between war and peace, an uneasy equilibrium that often seems on the brink of spinning out of control. As the pace of attacks rises, our vulnerability becomes more apparent each day: in the opening months of 2018, the federal government warned utilities that Russian hackers had put “implants” of malware in the nation’s nuclear plants and power grid and then, a few weeks later, added that they were infesting the routers that control the networks of small enterprises and even individual homes. In previous years there has been similar evidence about Iranian hackers inside financial institutions and Chinese hackers siphoning off millions of files detailing the most intimate details of the lives of Americans seeking security clearances. But figuring out a proportionate yet effective response has now stymied three American presidents. The problem is made harder by the fact that America’s offensive cyber prowess has so outpaced our defense that officials hesitate to strike back.
“That was our problem with the Russians,” James Clapper, President Obama’s director of national intelligence, told me one winter afternoon at a diner down the road from the CIA headquarters in McLean, Virginia. There were plenty of ideas about how to get back at Putin: unplug Russia from the world’s financial system; reveal Putin’s links to the oligarchs; make some of his own money—and there was plenty hidden around the world—disappear.
Yet, Clapper noted, “every time someone proposed a way to strike back at Putin for what he was doing in the election, someone else would come back and say, ‘What happens next? What if he gets into the voting system?’ ”
Clapper’s question drives to the heart of one of the cyberpower conundrums. The United States can’t figure out how to counter Russian attacks without incurring a great risk of escalation. The problem can be paralyzing. Russia’s meddling in the election encapsulates the challenge of dealing with this new form of short-of-war aggression. Large and small powers have gradually discovered what a perfect digital weapon looks like. It is as stealthy as it is effective. It leaves opponents uncertain about where the attack came from, and thus where to fire back. And we struggle to figure out the best form of deterrence. Is it better to threaten an overwhelming counterstrike? A non-cyber response, from economic sanctions to using a nuclear weapon? Or to so harden our defenses—a project that would take decades—that enemies give up attacking?
Naturally, the first temptation of Washington policy makers is to compare the problem to something more familiar: defending the country against nuclear weapons. But the nuclear comparison is faulty, and as the cyber expert James Lewis has pointed out, the false analogy has kept us from accurately understanding how cyber plays into the daily geopolitical conflict.
Nuclear arms were designed solely for fighting and winning an overwhelming victory. “Mutually assured destruction’’ deterred nuclear exchanges because both sides understood they could be utterly destroyed. Cyberweapons, in contrast, come in many subtle shades, ranging from the highly destructive to the psychologically manipulative.
Until recently, Americans were fixated on the most destructive class of cyberweapons, the ones that could turn off a nation’s power or interfere with its nuclear command-and-control systems. That is a risk, but the extreme scenario, and perhaps the easier to defend against. Far more common is the daily use of cyberweapons against civilian targets to achieve a more specific mission—neutralizing a petrochemical plant in Saudi Arabia, melting down a steel mill in Germany, paralyzing a city government’s computer systems in Atlanta or Kiev, or threatening to manipulate the outcome of elections in the United States, France, or Germany. Such “dialed down” cyberweapons are now used by nations every day, not to destroy an adversary but rather to frustrate it, slow it, undermine its institutions, and leave its citizens angry or confused. And the weapons are almost always employed just below the threshold that would lead to retaliation.
Rob Joyce, Trump’s cyber czar for the first fifteen months of the administration and the first occupant of that office to have once run American offensive cyber operations, described in late 2017 why the United States is particularly vulnerable to these kinds of operations, and why our vulnerabilities won’t go away anytime soon.
“So much of the fabric of our society rests on the bedrock of our IT,” said Joyce, who spent years running the Tailored Access Operations unit of the NSA, the elite operation charged with breaking into foreign computer networks. “We continue to digitize things; we store our wealth and treasure there; we run operations; we keep our secrets all in that cyber domain.” In short, we are inventing new vulnerabilities faster than we are eliminating old ones.
Rarely in human history has a new weapon been adapted with such speed, customized to fit so many different tasks, and exploited by so many nations to reshape their influence on global events without turning to outright war. Among the fastest adapters has been Putin’s Russia, which deserves credit as a master of the art form, though it is not the only practitioner. Moscow has shown the world how hybrid war works. The strategy is hardly a state secret: Valery Gerasimov, a Russian general, described the strategy in public, and then helped implement it in Ukraine, the country that has become a test-bed for techniques Russia later used against the United States and its allies. The Gerasimov doctrine combines old and new: Stalinist propaganda, magnified by the power of Twitter and Facebook, and backed up by brute force.
As the story told in this book makes clear, parts of the US government—and many other governments—saw all the signs that our chief adversaries were headed toward a new vector of attack. Yet the United States was remarkably slow to adapt to the new reality. We knew what the Russians had done in Estonia and Georgia a decade ago, the first time they used cyberattacks to help paralyze or confuse an opponent, and we saw what they later attempted from Ukraine to Europe, the testing grounds for cyberweapons of mass disruption and subtle influence. But an absence of imagination kept us from believing that the Russians would dare to leap the Atlantic and apply those same techniques to an election in the United States. And, like the Ukranians, we took months, even years, to figure out what hit us.
Worse yet, once we began to grasp what happened, a military and intelligence community that prides itself on planning for every contingency had no playbook of ready responses. In early 2018, when asked by the Senate Armed Services Committee how the National Security Agency and US Cyber Command were dealing with the most naked use of cyberpower against American democratic institutions, Adm. Michael S. Rogers, then coming to the end of his term as commander of both organizations, admitted that neither President Obama nor President Trump had given him the authority to respond.
Putin, Rogers said, “has clearly come to the conclusion that there’s little price to pay here and that therefore ‘I can continue this activity.’ ” Russia was not alone in reaching this conclusion. Indeed, many adversaries used cyberweapons precisely because they believed them to be a way of undercutting the United States without triggering a direct military response. North Korea paid little price for attacking Sony or robbing central banks. China paid no price for stealing the most private personal details of about 21 million Americans.
The message to our adversaries around the world is clear: cyberweapons, in all their various forms, are uniquely designed to hit America’s softest targets. And because they rarely leave smoking ruins, Washington remains befuddled about how to answer all but the biggest and most blatant attacks.
Rogers told me as he began the job in 2014 that his number-one priority was to “establish some cost” for using cyberweapons against America. “If we don’t change the dynamic here,” he added, “this is going to continue.” He left office, in 2018, with the nation facing a far larger problem than when he began.
In late July 1909, Wilbur and Orville Wright arrived in Washington to show off their Military Flyer. In the grainy pictures that have survived, Washington’s swamp creatures streamed across the bridges spanning the Potomac to see the show; even President William Howard Taft got into the act, though the Wright brothers were not about to take the risk of giving him a ride.
Not surprisingly, the army was fascinated by the potential of this wild invention. Generals imagined flying the craft over enemy lines, outflanking an oncoming force, and then sending the cavalry off to dispense with them. It wasn’t until three years later, in 1912, that someone thought of arming one of the new “observation aircraft” with a machine gun. Things both ramped up and spiraled down from there. A technology first imagined as a revolutionary means of transportation revolutionized war overnight. In 1913 there were fourteen military airplanes manufactured in the United States; five years later, with World War I raging, there were fourteen thousand.
And they were being used in ways the Wrights never imagined. The Red Baron shot down his first French aircraft in April 1916, over Verdun. Dogfights became monthly, then weekly, then daily events. By World War II, Japanese Zeros were bombing Pearl Harbor and performing kamikaze raids on my father’s destroyer in the Pacific. (They missed, twice.) Thirty-six years after Orville’s first flights in front of President Taft, the Enola Gay banked over Hiroshima and changed the face of warfare forever, combining the reach of airpower and the destructive force of the world’s newest ultimate weapon.
In the cyber world today, we are somewhere around World War I. A decade ago there were three or four nations with effective cyber forces; now there are more than thirty. The production curve of weapons produced over the past ten years roughly follows the trajectory of military aircraft. The new weapon has been fired, many times, even if its effects are disputed. As of this writing, in early 2018, the best estimates suggest there have been upward of two hundred known state-on-state cyberattacks over the past decade or so—a figure that describes only those that have become public.
And, as in World War I, this glimpse into the future has led nations to arm up, fast. The United States was among the first, building “Cyber Mission Forces,” as they call them—133 teams, totaling more than 6,000 troops, were up and running by the end of 2017. While this book deals largely with the “Seven Sisters” of cyber conflict—the United States, Russia, China, Britain, Iran, Israel, and North Korea—nations from Vietnam to Mexico are emulating the effort. Many have started at home by testing their cyber capabilities against dissidents and political challengers. But no modern military can live without cyber capabilities, just as no nation could imagine, after 1918, living without airpower. And now, as then, it is impossible to imagine fully how dramatically this invention will alter the exercise of national power.
In 1957, with the world on the nuclear precipice, a young Harvard scholar named Henry Kissinger wrote Nuclear Weapons and Foreign Policy. The book was an effort to explain to an anxious American public how the first use, a dozen years before, of a powerful new weapon whose implications we barely understood was fundamentally reordering power around the world.
One doesn’t have to endorse Kissinger’s conclusions in that book—especially his suggestion that the United States could fight and survive a limited nuclear war—to admire his understanding that after the invention of the Bomb, nothing could ever be the same. “A revolution cannot be mastered until it is understood,” he wrote. “The temptation is always to seek to integrate it into familiar doctrine: to deny a revolution is taking place.” It was time, he said, “to attempt an assessment of the technological revolution which we have witnessed in the past decade” and to understand how it affected everything we once thought we understood. The Cuban Missile Crisis erupted only five years later, the closest the world came in the Cold War to annihilation by miscalculation. That crisis was followed by the first efforts to control the spread of nuclear weapons before they dictated our fate.
While most nuclear analogies do not translate well to the new world of cyber conflict, this one does: We all live in a state of fear of how our digital dependencies can be hijacked by nations that in the past decade have discovered a new way to pursue old struggles. We have learned that cyberweapons, like nuclear weapons, are a great leveler. And we worry, with good reason, that within just a few years these weapons, merged with artificial intelligence, will act with such hyperspeed that escalatory attacks will take place before humans have the time—or good sense—to intervene. We keep digging for new technological solutions—bigger firewalls, better passwords, better detection systems—to build the equivalent of France’s Maginot Line. Adversaries do what Germany did: they keep finding ways around the wall.
Great powers and once-great powers, like China and Russia, are already thinking forward to a new era in which such walls pose no obstacle and cyber is used to win conflicts before they appear to start. They look at quantum computers and see a technology that could break any form of encryption and perhaps get into the command-and-control systems of America’s nuclear arsenal. They look at bots that could not only replicate real people on Twitter but paralyze early-warning satellites. From the NSA headquarters at Fort Meade to the national laboratories that once created the atomic bomb, American scientists and engineers are struggling to maintain a lead. The challenge is to think about how to defend a civilian infrastructure that the United States government does not control, and private networks where companies and American citizens often don’t want their government lurking—even for the purpose of defending them.
What’s missing in these debates, at least so far, is any serious effort to design a geopolitical solution in addition to a technological one. In my national security reporting for the New York Times, I’ve often been struck by the absence of the kind of grand strategic debates surrounding cyber that dominated the first nuclear age. Partly that is because there are so many more players than there were during the Cold War. Partly it is because the United States is so politically divided. Partly it is because cyberweapons were created by the US intelligence apparatus, instinctively secretive institutions that always err on the side of overclassification and often argue that public discussion of how we might want to use or control these weapons imperils their utility.
Some of that secrecy is understandable. Vulnerabilities in computers and networks—the kind that allowed the United States to slow Iran’s nuclear progress, peer inside North Korea, and trace Russia’s role in the 2016 election—are fleeting. But there is a price for secrecy, and the United States has begun to pay that price. It is impossible to begin to negotiate norms of behavior in cyberspace until we too are willing to declare our capabilities and live within some limits. The United States, for example, would never support rules that banned cyber espionage. But it has also resisted rules prohibiting the placement of “implants” in foreign computer networks, which we also use in case the United States needs a way to bring those networks down. Yet we are horrified when we find Russian or Chinese implants in our power grid or our cell-phone systems.
“The key issue, in my opinion,” says Jack Goldsmith, a Harvard law professor who served in George W. Bush’s Justice Department, “is the US government’s failure to look in the mirror.”
On a summer day in 2017, I went to Connecticut to see Kissinger, who was then ninety-four, and asked him how this new age compared to what he grappled with in the Cold War. “It is far more complex,” he said. “And over the long-term, it may be far more dangerous.”
This book tells the story of how that complexity and danger are already reshaping our world, and explores whether we can remain masters of our own invention.