New Cold Wars: China's Rise, Russia's Invasion, and America's Struggle to Defend the West

by David E. Sanger

This event sparked cries of the infamous “missile gap” bolstered by government charts warning that the Soviet missiles were superior in number and quality. The missile gap turned out to be pure fiction, as the Kennedy administration itself was forced to concede once in office, but by then the facts didn’t matter: The American shock over the Soviet achievement triggered the space race, catalyzed the creation of NASA, and ultimately led to the fulfillment of Kennedy’s goal to land a human on the moon by the end of the decade.

By 2018, the Chinese payoff was already becoming apparent. In August, they claimed to have successfully tested the Xingkong-2 “Waverider”—a hypersonic cruise missile capable of moving forward on its own shockwaves. (The U.S. Air Force had been developing a similar concept called the X-51 but allowed the program to die out in 2013.) In 2019, China claimed it had added the “Dongfeng-17” to its arsenal—a hypersonic glide vehicle, similar to what was ultimately part of the “Sputnik moment” test—mounted on a rocket so powerful that it was dubbed the “aircraft-carrier killer.” Also debuting in 2019 was the WZ-8: an uncrewed spy plane with the potential to carry a nuclear warhead.

The alarmist view of the Chinese and Russian buildup was that America needed a crash program to match the progress being made by its adversaries. That was the essence of Milley’s comparison to Sputnik: that this is a race we have to win, no matter the cost. But while that might have been possible in 1960, Sputnik moments looked different in the 2020s. There were so many technologies that the United States was competing for advantage in—hypersonics and space weapons, semiconductors and artificial intelligence, cyberdefenses and biodefenses—that you could have a Sputnik moment just about every month. And Washington simply wasn’t prepared to sort out the vital from the merely important, and the technologies that America must have from those it could survive without.

It was also true that the current standoff between the United States and China was fundamentally different at its core: The deep links between the two economies—the mutual dependencies on technology, trade, and information that leaps the Pacific in milliseconds on American- and Chinese-dominated networks—never existed in the old Cold War. The Berlin Wall not only delineated a sharp line between spheres of influence, freedom, and authoritarian control, it stopped most communications and trade. By contrast, in 2021, China was the largest supplier of goods to the United States and the third-largest consumer of its exports, after Canada and Mexico. That was a huge difference from the Cold War era, when, beyond caviar and vodka, the United States had no real interest in Soviet exports.

Of all the military developments in China that American officials had worried about over the previous few years—cruise missiles and satellite killers and threatening sorties around Taiwan—new nuclear weapons had not ranked among them. Mao Zedong had always taken a moral-high-ground pride in designing China’s “minimum deterrent” strategy, which held that a few hundred nuclear weapons was all Beijing needed to deter an attack from larger powers. While the United States had briefly considered attacking China’s nuclear test facilities in 1964 (perhaps with Soviet help), in the end it decided to learn to live with the Chinese Bomb. After all, even by the double standards of the West’s nuclear calculus—in which Western powers with nuclear weapons are fine and everyone else a mortal threat—it seemed hard to complain about a few hundred Chinese nuclear weapons when the United States and Russia both maintained deployed arsenals five times larger, not including thousands more weapons that Washington and Moscow kept in storage.

For decades intelligence collectors thought they had one job: Steal or gather secrets and hand them off to the president or other officials to give them an advantage. Secrecy was baked into the cake. Now they were being told that secretly collected information might be most useful if it is made public. It was a huge shift. “What,” Nakasone asked me rhetorically, “are the cultural implications of a place like NSA releasing information like that? You can imagine that there are people who have spent their livelihood collecting this information.” This was the first time they were being asked—actually, ordered—to release it publicly.

Pompeo, who was fresh off a term as the CIA director, about this, he told me the two countries had such long-running enmity that they would never work together effectively. In fact, the history between the two countries wasn’t so clear-cut: Like planets moving in and out of each other’s orbits, there were moments when they moved closer and farther apart. But this history—when they worked together and when they blasted apart—illuminated what made their cooperation in 2022 so different.

So it was no surprise that Mao began seeking his own atomic bombs—aiming to make sure that he would never be dependent on the Soviet arsenal. Naturally, a nuclear-armed China posed a threat to America, but it also posed a threat to the Soviet Union. The Johnson administration thought about inviting the Soviets to jointly strike Lop Nor, the Chinese nuclear testing site. William C. Foster, the director of the Arms Control and Disarmament Agency, said in one recently declassified memo that “if we could get together with the USSR, the Chinese could be handled even if it required an accidental drop on their facilities.” Wisely, the Johnson administration abandoned the idea, determining in April 1964 that the risk of a Chinese nuclear capability “is not such as to justify the undertaking of actions which would involve great political costs or high military risks.”

What followed was America’s unipower moment. China used the coming years to lift more people out of poverty than any World Bank program ever could. For its part, Russia used it to create oligarchs and rearm but failed at the key task of diversifying its economy. And it barely noticed as a generation of innovators, starting with Sergey Brin, one of the founders of Google, took their talents to the United States.

As Pentagon officials watched the invasions unfold, they were also struck by the evidence that Russian supply and logistics operations were hopelessly snarled and backlogged. Not only had the Russians failed to bring along enough food to sustain a battle of more than a few days, but the column of Russian troops marching down to Kyiv had stalled out entirely. The forty-mile-long traffic jam of Russian tanks, personnel carriers, and fuel tanks became the iconic image of early Russian blundering.

It wasn’t, of course, just the Ukrainians who felt the loss of Viasat. Tens of thousands of customers around Europe and North Africa were disconnected too—everyone from broadband companies in the U.K. to utilities in Germany. But by disrupting key command and control for the Ukrainian forces in the opening hours of a major conflict, the stakes were radically higher: Without communications, the military couldn’t direct its soldiers, share information, coordinate among different branches, or share warnings and updates. They likely had other, more limited means to talk to each other—but not on the scale that was needed to repel an invasion. It might have worked, too. But early on the morning of February 26, Mykhailo Fedorov, Ukraine’s deputy prime minister for digital transformation, sent a plea to the one person he was confident could get Ukraine back online: Elon Musk, for a time the world’s richest man. Musk’s company SpaceX had developed a constellation of low earth orbit satellites called Starlink that promised its users the ability to get online almost anywhere in the world. Armed with a Starlink terminal, users could tap into the network of thousands of tiny satellites operating high above.

There were two other factors that likely shaped how the cyber war played out. The first was that cyber has long been the right tool for “short-of-war” conflicts, when the objective is to undercut or influence an adversary without triggering a full-on war. But when the shooting starts, the utility of modern cyberweapons can wane. There is often not enough time, or enough access to networks, to use it effectively. Why spend months seeking the perfect entrée into the electric power grid, and figuring out what code might shut it down, when you can aim a missile directly at the power plant?

lifelong diplomat, she had served for a time in Moscow. When we spoke, she was just a few months short of retirement, perhaps making her more willing than most serving diplomats to assess what she and her colleagues had missed. At its core, she concluded, diplomacy was premised on “a wager, a bet on a result, or a desire. We end up being right or wrong. And in the end”—on Putin, on Russia’s direction for the future—“we got it wrong.”

What happened next is still in some dispute. The Navy provided a general confirmation but no exact coordinates. That was for a reason: The rules that the United States established prohibited giving certain types of data that led to the deaths of Russian troops or sailors. Legally, that would have made the United States a more direct combatant. Still, the fact was that the United States was getting deeper and deeper into the conflict, walking right up to the line between combatant and noncombatant.

The key, national security officials argued, was that the Ukrainians were the only ones making the final decision about whether, where, and when to shoot. Americans, the White House and Pentagon lawyers said, could collaborate on devising Ukrainian military strategy, arm Ukraine’s soldiers, and tell them generally where the enemy would be and when they would be there—all while avoiding becoming a “co-combatant” in the conflict. It was a finely turned legal distinction, but one very important to Biden, Jake Sullivan, and Antony Blinken, who had to make the case that the United States was aiding Ukraine, not fighting Russia.

But even if that happened, Jake Sullivan was quick to point out, Ukraine would be under constant threat for years, maybe decades—a threat so omnipresent that it would need to be able to deter Russia from another invasion, whenever Putin rebuilt his sorry force. Meeting that challenge would require an increase in aid and support on a scale that NATO, that Congress, and that even the Ukrainians had never thought about before. “When you think about what we provided in 2021, it was more than we had provided ever before,” Sullivan pointed out much later, looking back at the early days of the war. “It was less than a billion dollars.” That amount was tiny compared to the kind of numbers now kicking around the Pentagon. For Ukraine to survive over the long term, its military would need to be completely overhauled. It needed to become like Israel, said one former military official who was now serving in the Biden administration. It would have to go from a force that was dependent on decrepit Soviet-era leftovers to modern, Western arms—all while fighting a brutal war in real time on its own territory. It might not be a member of NATO for a long, long time. But it needed to be armed like one.

How should the aid be paced, given the changing nature of the war and the limits on getting artillery shells into a long supply chain that stretched from the Army Ammunition Plant in Scranton, Pennsylvania—just three miles from Biden’s childhood home—to the border towns of Poland and then into Ukraine? How much risk were they willing to take when it came to taunting Vladimir Putin? How much blame—and domestic public pressure—were they willing to take for refusing Zelensky’s requests? None of these were easy to answer in the fog of a fast-moving war. But over time, a strategy emerged that some in the administration began to call “boiling the frog.” The phrase is based on the myth that a live frog, placed in a pot and slowly boiled, will not leap out if the temperature rises incrementally enough.

2023, a Times team determined that former Ukrainian special forces operators were behind the explosion, and in summer 2023 The Washington Post broke a fuller story: Six covert Ukrainian operatives had rented a sailboat and placed the explosives, in a mission likely approved by Zaluzhny and hidden from Zelensky to give him room to deny Ukrainian involvement.

It was a wonderful non-compromise. Su told me that for a number of years he believed that Taiwan might truly be able to “float” in between the great powers, supplying both, siding with neither. As their relative power rose and fell, Taiwan could shift alliances and strategies accordingly—a sort of “strategic equilibrium” that others have attempted. It was very French. Naturally, it couldn’t last. Over the next few decades, the delicate political balance—held together chiefly by mutual economic benefit—began to fray at the edges. While trade increased, China used every bit of its economic leverage to force countries to abandon their diplomatic recognition of Taiwan and embrace the mainland as the only true China. Only Vatican City and about a dozen small powers recognized Taiwan, and Beijing was picking them off one by one.

breaking point came as Xi Jinping consolidated power. Gradually, it became clear that all the American intelligence analysis stating that he would focus on the economy and avoid conflict was wrong. Xi was determined to control the outliers—starting with Tibet, Hong Kong, and the Muslim-majority area of Xinjiang. When the last of Hong Kong’s street protesters were finally crushed in 2020, it was clear that Taiwan could easily be next.

So “Atoms for Peace” was based on an explicit deal: Countries that forswore nuclear weapons would get help from the United States and the other nuclear powers to build commercial nuclear power plants. The lure of seemingly limitless sources of electricity, the United States was betting, would outweigh the desire for a nuclear weapons option. It was the job of the IAEA, created in 1957, to make sure that the plants were safe and that no nuclear fuel was surreptitiously diverted. Its mandate is deliberately narrow: Contrary to popular belief, IAEA inspectors don’t search for weapons, just the fuel that powers them. In fact, there are only a handful of true weapons experts in the tall towers in Vienna overlooking the Danube River.

Grossi initially believed that Iran, along with North Korea, would dominate his next few years. Donald Trump’s decision in 2018 to pull out of the Obama-era agreement that sharply constrained Iran’s nuclear fuel production was coming around to bite the rest of the world: Iran was resuming production and limiting inspections, despite the best efforts of the Europeans to keep the deal alive.

He marveled that the Russians had sent their troops through the Chernobyl Exclusion Zone—where the soil was still radioactive—without protective gear. Many slept in the forest.

The Ukrainian leadership all but dismissed the nuclear threat. In their view, Putin would threaten repeatedly but never press the button. Even if he did, some told me, how could it be worse than his leveling of whole cities? Their bigger fear was that Putin’s bluff would work and lead the West to back off. This was the unspoken follow-on to Milley’s nuclear paradox: The more the Ukrainians were successful and Putin blustered, the more tempted the West would be to warn the Ukrainians against taking back too much of their own territory.

There was the moment, in 1979, when an American nuclear watch officer reported that sensors were warning of a massive, incoming Soviet strike. It turned out that someone had loaded a training tape into the real warning system. In September 1983, a Soviet lieutenant colonel saw a warning flashing that five American intercontinental ballistic missiles were coming straight for the Soviet Union. Trusting his gut, he called the Soviet army headquarters to report a system malfunction. Just months later, nuclear calamity nearly struck again when a 1983 NATO training exercise designed to culminate in a simulated nuclear attack got a little too real—and the Soviets began to prepare for war.

if the first lesson of 1962 had been that it was the height of folly to make an adversary choose between humiliation and the nuclear option, the lessons that followed were that human error, digital malfunction, miscommunication, or sheer inattention to detail could be equally deadly.

When the new German chancellor, Olaf Scholz, was on his way to Beijing to meet Xi Jinping, one of his aides told me, a condition of the meeting was a public statement from the Chinese leader that there was no role for nuclear weapons in the Ukraine war. On November 4, with Scholz by his side, Xi “reaffirmed” that he was in favor of a negotiated settlement to the crisis. It was critical, state media wrote—quoting their leader—that everyone involved in the crisis “oppose the use of or the threat to use nuclear weapons, advocate that nuclear weapons cannot be used and that nuclear wars must not be fought.” To Biden’s mind, those statements made a difference. Once Putin’s most important protector said nuclear use had to be off the table, the Russian leader backed off, if grudgingly. At a meeting of Russia’s Human Rights Council at the Kremlin on December 7, he repeated his familiar line that Russia would use “all available means at our disposal” to defend itself. But then he went on, in a statement that seemed directed at Xi and others who thought he had pushed too far. “We have not gone crazy…. We are not going to brandish these weapons like a razor, running around the world.”

TSMC only makes one thing: computer chips. While it produces all varieties of high-quality semiconductors—from memory and graphics chips to the kind of advanced, custom-designed microprocessors that power Apple’s iPhone—it made its name by focusing purely on precision, low-cost manufacturing in the fabrication plants (or “fabs”) it built in the Science Park and then scattered across the island. It doesn’t design chips. It takes other company’s designs and churns out the chips the world relies on, for everything from F-35s and HIMARS artillery systems to Android cellphones and smart TVs. Today politicians regularly call chips “the new oil”—the analogy is flawed, but it does capture their centrality to the global economy.

Chip manufacturing is the technology America let get away. Silicon Valley’s early pioneers at Fairchild Semiconductor and Texas Instruments, Intel and Motorola had pioneered chip technology and changed the world. But over the decades, those companies and others, in a series of separate profit-driven decisions, gradually moved production overseas without much thought about the national security implications of becoming so dependent on the supply of chips from a single vulnerable island off the coast of China. It was assumed that, like oil through the Strait of Hormuz, the chips would just keep flowing across continents and borders.

TAIWAN’S EMERGENCE AS the critical hub of the most important supply chain in the world was the result of a remarkable confluence of events: unusually skillful and well-timed state-planning by Taiwan at a moment when it was barely emerging as a democracy; remarkable vision by a single brilliant engineer, Morris Chang, a Harvard-and-MIT-educated U.S. immigrant who left Texas Instruments in disgust when he felt underappreciated; and a series of astoundingly bad business judgments made in the United

development, K. T. Li, to come for a visit—and a little arm-twisting. Among Taiwan’s leaders, Li best understood that the island was stuck at the low-profit end of the semiconductor supply chain—mostly packaging chips or integrating them into cheap electronics. If it wanted to move up the ladder, to prosper like Japan, it needed a new business plan that would put it at the heart of the manufacturing process.

Washington was accustomed to reviewing Chinese investments in American tech firms and stopping them whenever they detected a threat to national security. There was no equivalent process to stopping companies from moving their production overseas—that was purely a business decision. So over the years, American tech firms decided to focus on designing their own chips and leave the manufacturing to the true experts like TSMC. It was a strategy reinforced by the reassuring myth that in a globalized world, it didn’t really make any difference where you produced the semiconductors that fuel the information age. Sure, there might be temporary disruptions, but in a region where making money was the supreme goal, keeping supply chains flowing seemed in everyone’s economic interest—especially China’s.

One rule blocked the sale to China of the most advanced chips—particularly those suited for artificial intelligence applications—and the U.S.-built chipmaking manufacturing equipment, materials, and software that China would need to try to make its own alternatives. The second made it simpler for the Commerce Department to ban trade with specific companies or individuals.

American companies were not exactly thrilled, either. As much as they wanted Biden’s administration to safeguard their intellectual property and subsidize new manufacturing sites, they did not want to be cut off from lucrative revenue streams in China. Chief among them was Nvidia, the onetime game maker whose A100 chips had emerged as the fastest processors for assembling the “large language models” needed for artificial intelligence systems. The company was quickly realizing these export controls could cost it billions in lost sales. For Nvidia there was one obvious solution—it immediately designed two “China market” chips that fell just below the speed limits established by the Commerce Department. Fairly quickly, White House officials were calling Situation Room meetings on how to crack down on the chipmaker, tightening the rules so that Chinese users couldn’t string several Nvidia chips together. It was clear that with every new rule there would be new workarounds.

The overriding challenge in chipmaking has always been to pack more and more tiny transistors on each successive generation of semiconductors. More circuits translated into more and faster computing power. For decades, there was a rough rule for the pace of innovation in the industry: every two years, the number of transistors on a chip roughly doubled—that was the essence of Moore’s law, named for the same Intel co-founder who coined the easy metric of progress.

eventually, the laws of physics began to intervene. The cost of shrinking the circuitry skyrocketed, and miniaturization slowed. Making breakthroughs became a multi-billion-dollar enterprise, requiring ever thinner beams of light, moving in smaller and smaller wavelengths, to produce ever smaller circuits. And ASML’s machines, with their special lens coatings and innovative light sources, became the only ones that could effectively produce these types of chips at scale. By the time Rutte came to the White House to see Biden, the firm’s machines could produce chips with circuits that are only three nanometers wide: roughly thirty thousand times thinner than a human hair.

machines, they will develop them themselves.” There were already hints that the Chinese were cracking the code. In the weeks just before the CHIPS Act passed, SMIC produced a seven-nanometer chip that was used in a Chinese system for mining cryptocurrency. It was a surprise: As recently as 2020, amid the pandemic, China had appeared stuck at forty nanometers. The CIA spent weeks examining the chip and concluded it was based on, or perhaps stolen from, TSMC’s own designs. A year later, in the late summer of 2023, Huawei brought out a new cellphone using a chip with the same circuit dimensions. The message seemed clear: The United States could deprive Beijing of the tools, and impose embargo after embargo. But China was determined to find its own way to the goal.

“Competition with China has begun to consume U.S. foreign policy,”

“The instinct to counter every Chinese initiative, project, and provocation remains predominant, crowding out efforts to revitalize an inclusive international system that would protect U.S. interests and values even as global power shifts and evolves.”

Just before Campbell arrived, Prime Minister Manasseh Sogavare announced that he’d signed a new security pact with Beijing, the contents of which were being kept secret. But there were hints: Leaked documents suggested that in exchange for showering Sogavare’s government with economic aid, China would be empowered to bring in its armed forces and military police to protect its own interests—or Sogavare’s political survival. The bottom line was that Beijing gained another military foothold in the region, within 1,200 miles of Australia.

A year after the fiftieth anniversary commemoration of the battle of Guadalcanal—a days-long celebration in which American veterans returned to the beaches and told stories of a desperate battle—the State Department closed the embassy on the Solomons, hoping to save, as Farrand recounted, $387,000 a year. Over the following decades, the United States constantly made promises—to reopen the embassy, to send in the Peace Corps, and to invest in a deepwater port. But there was no follow-through. Aid dribbled in—for disaster relief, to remove mines, to gain United States access to fisheries—but the Solomons were not an American priority.

Few were more bitter about that absence than Sogavare, who began his fourth term as the country’s prime minister in 2019. He had gradually come to the conclusion that the Americans talked a big game, but the Chinese signed big checks. And as the paranoid, corrupt, and increasingly autocratic leader of a perpetually broke country, Sogavare considered those checks very important.

There was a cost, of course. In 2019, Beijing made it clear that it had one non-negotiable demand: The Solomons had to cut all diplomatic ties with Taiwan and recognize the mainland. Sogavare acquiesced, despite the riots that followed. (Taiwan had long offered free healthcare to islanders as well as substantial development and educational funding; there were suspicions of bribery surrounding the diplomatic shift.) It turned out that the quid pro quo for abandoning recognition of Taiwan was $730 million in Chinese aid.

China’s investment in the South Pacific was long in the making. Back in 2013, when Xi announced his signature development and trade undertaking, the Belt and Road Initiative, the Chinese said the loosely centralized program was a source of “win/win cooperation.” Recipient states would get quick and easy funding for new, often sorely needed infrastructure projects—bridges, hospitals, roads, ports—without having to comply with the onerous requirements that Western-led institutions often demanded. Yet when American diplomats saw the deals, they winced at how efficiently Beijing was leveraging its coffers to gain profit and influence. The funding might have been easy, but it was structured in a way that saddled countries with debt they would struggle to pay back: Interest rates from Chinese creditors were often four times higher than the Western equivalents, researchers found, and had to be repaid in a third of the time,

Frequently written into the contract was a stipulation that if the country fell behind on debt repayment, the project—typically, domestic infrastructure in strategic locations—or other key concessions would transfer to China. Beijing dismissed these fears as no more than a Western “smear campaign.”

port was handed over to a Chinese company with a ninety-nine-year lease. The simple explanation was that an unprofitable port had been leased out to a foreign company for a profit. The more pernicious alternative was that Beijing, using a tactic known as “debt-trap diplomacy,” was putting itself in a position around the world to take over strategic infrastructure.

But at the same time, Niinistö was also meeting regularly with Putin, taking his measure. It didn’t take long for Putin to make it clear that in his view what was his was his—and what used to be Finland’s was also his. “In our first meeting—that was 2012—he suddenly said to me, ‘Why are you heading for NATO? You can’t get Karelia back,’ ” said Niinistö, referring to territory that the Soviets took from Finland in the last days of World War II. (The Karelian region has changed hands many times over the centuries; today it is divided between Finland and Russia.)

During the Cold War, this careful wending between Moscow and the West came to be known as “Finlandization,” a somewhat pejorative shorthand for the deal with the devil that the country struck to avoid Russia’s wrath and retain its own independence. By steering clear of formal alliances—and not allowing NATO forces to sit on its unfenced eight-hundred-mile border with Russia—Finland could avoid contributing to Putin’s paranoia that the Western alliance

was creeping east and surrounding him. Although the need for Finlandization ended with the Cold War, and the country later joined the European Union, it steadfastly refrained from joining NATO—largely for reasons of regional stability, it argued, and always retaining the right to join if the situation changed. While Finland didn’t have the formal protection of NATO membership, it also had no obligation to go to war with Russia if any NATO member suddenly found itself the victim of a Russian onslaught. Or at least, that is what Niinistö could say to Putin. Whether Putin believed it was another matter.

In the year since the invasion, working the phones and traveling to Washington every few months, Niinistö moved his country firmly into the Western camp. He applied for full NATO membership, along with Sweden, a fact that Biden mentioned time and again, always to twist the knife that Putin had accomplished “exactly what he had tried to avoid”: an expansion of NATO. What Biden blew past, of course, was how hard that path had been. Even in the urgency of the war, NATO hadn’t been able to get out of its own way. What seemed like a clean vote to admit two qualified members—Finland and Sweden—got hung up, as Turkey spent the last several months withholding its vote in an effort to leverage political concessions from the Swedish government.