Credit...Photo illustration by Pablo Delcan

How TikTok Became a Diplomatic Crisis

A Chinese app conquered the planet — and now the U.S. is threatening to shut it down. Can the world’s biggest virality machine survive?

Listen to This Article

Audio Recording by Audm

To hear more audio stories from publications like The New York Times, download Audm for iPhone or Android.

On March 10, two weeks after Russia’s invasion of Ukraine, the White House convened a Zoom call with 30 prominent TikTok creators. Jen Psaki, then the White House press secretary, and members of the National Security Council staff briefed the creators, who together had tens of millions of followers, on the latest news from the conflict and the White House’s goals and priorities. The meeting followed a similar effort the previous summer, in which the White House recruited dozens of TikTokers to help encourage young people to get vaccinated against Covid.

The app had only become more popular in the intervening months. “We recognize this is a critically important avenue in the way the American public is finding out about the latest,” the White House director of digital strategy, Rob Flaherty, told the assembled group. “So we wanted to make sure you had the latest information from an authoritative source.” Yet at the same time, the Biden administration was more than a year into negotiations with ByteDance, the Chinese company that created and owns TikTok, about national security concerns surrounding the app. In fact, the White House staff members who organized and briefed the TikTok creators were barred from downloading the app on their work phones.

The administration’s contradictory approach to TikTok — its embrace of the app as a vital conduit to the public, and its fear of the app as a potential tool of foreign influence — is perhaps a fitting response to the utterly unique problem that TikTok poses. Seemingly overnight, TikTok has managed to remake American culture both low and high, from media and music to memes and celebrity, in its own image. TikTok turned Olivia Rodrigo into a household name and propelled the author Colleen Hoover to the top of the best-seller list, with more copies sold this year than the Bible. TikTok coined “quiet quitting,” one of the hallmark phrases of 2022, and introduced a whole new dialect of algospeak — “seggs,” “unalive,” “le dollar bean” — that is now spreading across pop culture. Corporations and brands, from Goldfish crackers to Prada, have redirected billions of dollars worth of advertising to the platform in recognition of its all-encompassing reach, which can, at seemingly any moment, turn even a decades-old product into a must-have item. Last year, TikTok had more site visits than Google, and more watch minutes in the United States than YouTube. Facebook took almost nine years to reach one billion users; TikTok did it in five.

The app’s extraordinary success is made even more remarkable by the fact that it is a product of America’s greatest geopolitical rival. Despite decades of trying, no Chinese company has ever conquered American society like TikTok. It’s difficult to imagine a Russian or Iranian company — or, increasingly, even another Chinese company — pulling off a similar feat. TikTok’s provenance has stoked persistent and longstanding worries about its vulnerability to exploitation and manipulation by the Chinese government. Over the last year in particular, TikTok has faced an unceasing stream of bad press, with each week seeming to bring a fresh revelation about the company’s questionable data practices and spotty internal safeguards. In just the last six months, TikTok and ByteDance have been accused of lying about the access of China-based employees to American user data, using a news app to push pro-Beijing content abroad and allowing Chinese state media accounts to run unchecked and unlabeled as they criticized the American political process.

If TikTok’s popularity has thus far provided it some insulation against government action, the app’s time may be running out. In November, Brendan Carr, a commissioner of the F.C.C., said it should be banned outright. Senator Mark Warner, co-chairman of the Senate Select Committee on Intelligence, said of a ban, “The sooner we bite the bullet, the better.” Christopher Wray, director of the F.B.I., told Congress he was “extremely concerned” about TikTok’s operations in the United States. Earlier this month, Senator Marco Rubio introduced legislation that would effectively prevent TikTok from operating in the United States by banning all apps “subject to substantial influence” by China, Russia and other foreign adversaries.

The Biden administration, meanwhile, is said to be nearing a deal with TikTok. Speaking at a conference of tech leaders, Kemba Walden, the principal deputy national cyber director, said that the White House had not made any final decision on a ban, but voiced support for “any measure that will raise security.” Maryland, South Dakota, South Carolina, Nebraska, Texas, Alabama and Utah have already banned use of the app on state devices. A bill passed by the U.S. Senate last week would do the same at the federal level. The military has also barred the app from government devices.

What often goes unnoticed in these conversations is that TikTok is as much a product of the West as it is of China. ByteDance owes its very existence to the intermingling of ideas, capital and people that defined the last five decades of U.S.-China engagement. The United States sought to woo China with the appeal of its model and the benefits of the existing international order, in the hope that a liberalized market economy would foster domestic political reform. At the same time, Beijing seemed eager to build up its own tech sector as an engine of economic growth and global soft power. The success of a product like TikTok was only the most visible example of a deeper tech symbiosis that once appeared inevitable.

But now the world has changed. In the United States, being tough on China is one of the few areas of bipartisan agreement. And in this fraught geopolitical context, TikTok is considered a Trojan horse — for Chinese influence, for spying, or possibly both. In China, meanwhile, a broad crackdown has sought to rein in high-flying tech companies and their founders, out of fear that, with their influence, independence and popularity, they were becoming alternative power bases to the Chinese Communist Party. The campaign is only one part of a broader political and social chill that threatens to pull the country back to the days of Mao. TikTok itself is not available in China — users there must access a different ByteDance app, which follows Chinese government directives on censorship and propaganda.

Just a few years ago, the rise of ByteDance seemed like a harbinger of an era of Chinese app dominance. Indeed, it would be hard to find a company more self-consciously modeled, in both spirit and substance, on America’s tech giants. ByteDance’s founder internalized the mythos of Silicon Valley, taking to heart the idea, long promoted by Washington, that the American market was the ultimate prize, and that it welcomed any entrepreneur with the talent and ambition to succeed. But now, with walls going up on both sides of the Pacific, TikTok seems likely to be the last of its kind as well as the first. The company is caught in the middle between the old era and the new — too Chinese for America, too American for China.

Image
Credit...Photo illustration by Pablo Delcan

On Christmas Day in 2010, a short, bespectacled 27-year-old Chinese programmer named Zhang Yiming logged onto Douban, a Chinese hybrid of Rotten Tomatoes and Goodreads, to share his thoughts on a movie he had just watched. Zhang used his Douban account as a chronicle of his personal development, recording the books he wanted to read (“What Would Google Do?” “Catch-22” and “The Road to Serfdom”) and the movies he’d seen (“The Departed,” “Good Will Hunting,” “Inception”). The movie Zhang watched that Christmas was “The Social Network.” The movie was of particular interest to Zhang; he was only a year older than Mark Zuckerberg and, after several years spent bouncing between small start-ups and an unhappy stint at Microsoft, he had recently become a founder himself. Zhang’s company, 99fang, was a real estate search engine, but he had ambitions to build something bigger. The story of Zuckerberg and his ruthless climb to the top was both inspiration and warning. He gave it four out of five stars.

Born in 1983 as the only son of a librarian and a nurse, Zhang came of age in a China flush with reform and newfound connections to the West. He moved to Tianjin for college, where he studied computer engineering. Zhang loved the freedom that technology offered and displayed a fondness for the West, politically as well as culturally. In 2009, when Chinese authorities blocked access to several websites, he took to his personal blog to voice his disapproval, according to a Wall Street Journal profile. “Go out and wear a T-shirt supporting Google,” he wrote. “If you block the internet, I’ll write what I want to say on my clothes.”

In 2011, smartphone shipments in China exceeded those in the United States for the first time. Riding the subway in Beijing, Zhang noticed that fewer and fewer people were reading newspapers; instead, they turned to their phones to pass the time. Zhang, a restless entrepreneur, developed an idea for a new company, one that would take advantage of the rise of the mobile internet and the birth of early artificial intelligence. “Just like Zuckerberg founded Facebook to connect people with people, and Travis founded Uber to connect people with cars,” he later said, he wanted to “connect people with information.”

The problem with a lot of the internet, Zhang believed, was the paralyzing number of options it presented to users. The website was “an outdated form of information organization,” he told an interviewer — messy and imprecise and filled with extraneous information. And the RSS feed was an unsuitable replacement, he wrote in a now-deleted blog post, because people were “forced to figure out ‘what I like and what I want’ themselves.” At one meeting with an investor, Zhang sketched out on a napkin his vision of a superior system, powered by artificial intelligence — it would be information finding users, rather than the other way around. In early 2012, he started ByteDance.

An early backer took Zhang to almost two dozen Chinese investors, but none of them were interested. Digital media was not a great business, and China had its own emerging internet giants that could simply copy anything novel ByteDance devised. It didn’t help that Zhang was still so baby-faced that, when introduced, people often ignored him and spoke to his colleagues instead. On top of this, Zhang was not an A.I. expert, nor were any of ByteDance’s early employees. There were not yet any detailed reference texts available in Chinese, and, according to Chinese Entrepreneur Magazine, when Zhang tried to buy one that was in pre-publication, he was rebuffed. Zhang eventually taught himself to write a recommendation engine from scattered resources on the internet.

A few months after ByteDance’s founding, Matt Huang, an American investor and venture capitalist, visited Beijing to check out the tech scene. A friend told him about ByteDance and set up a meeting. When Huang arrived at the company’s offices, he found a humble operation, with Zhang overseeing about 20 people jammed into two apartments. After two hours of conversation, Huang was ready to invest in Zhang’s idea, despite hesitations about the business model. “I was skeptical of the idea but blown away by the person,” Huang said. Most of his lingering doubt came down to the viability of applying A.I. to a news app. But, he said, “I thought if someone was going to figure it out, it was him.”

In August 2012, ByteDance launched the first version of Jinri Toutiao (“Today’s Headlines”), an app that used A.I. to, in Zhang’s words, “let every user, at every moment, see their own front news page.” Toutiao pioneered the system that TikTok would later ride to global dominance. Other content platforms, like Facebook and Twitter, required users to manually accumulate friends and connections, whose posts then populated the user’s feed. Toutiao, by contrast, didn’t care whom you knew, only what you liked. It required no sign up at all — no need to create an account and password, or to describe interests or preferences. Users were simply presented with articles upon downloading the app. Based on how a user reacted to a piece of content — reading the whole article or just a few sentences, pausing on a particular paragraph, swiping back up to read something again, leaving a comment — Toutiao’s underlying technology began to generate a picture of who the user was and what they wanted.

Best of all, the recommendation engine got better with every use. It was a virtuous cycle: more users meant more data; more data meant a smarter algorithm; a smarter algorithm meant more users; and on and on. The app hit one million daily average users only four months after it started. By one count, revenue grew from almost nothing in 2014 to $2.5 billion in 2017. The bigger it got, the more criticism Toutiao faced. Detractors said the app catered to the basest human interests — celebrity gossip, scandal, disaster and violence — to keep users hooked as long as possible. The average user was spending more than an hour a day on the app, and Toutiao was, by revenue, one of the fastest-growing apps in the history of the internet. Still, Zhang could see the ceiling: According to the Chinese business press, an internal company assessment placed the total size of China’s newsfeed market at around 240 million daily average users. If Toutiao claimed half the pie, it would max out at 120 million daily average users. To keep growing, the company had to look abroad.

But the push also reflected something more fundamental than business strategy. Chinese companies, Zhang has said, are “born to be global,” just like American companies. “There was a different kind of leadership, a hunger for growth,” a former ByteDance executive told me. “It was a challenge, in some sense self-imposed — if the big Western tech can do it, why can’t we? We are no less.”

In 2014, Zhang visited Silicon Valley with a group of Chinese founders, touring the offices of Facebook, Tesla and Airbnb. He spotted Xiaomi phones in the hands of American tech workers, and heard conversations about Alibaba’s hotly anticipated American I.P.O. Back in Beijing, Zhang summarized his feelings on his blog: “The golden age of Chinese tech companies is coming.”

A lip-syncing app called Musical.ly was already showing the way. Started in Shanghai by a pair of Chinese entrepreneurs, Musical.ly was a surprise hit in the U.S. market, particularly among teenagers. Musical.ly’s rise intrigued ByteDance, which was on the hunt for a new product to plug into its recommendation engine. In March 2016, the company assigned a handful of employees to a new initiative called Project X, with the goal of replicating Musical.ly as closely as possible. The team discovered that none of the existing Chinese video-sharing apps — there were hundreds of them — had the technology to sync videos with soundtracks. The gap, of about 200 to 300 milliseconds, was small, but noticeable enough to turn off potential users. Fixing the syncing issue was the original selling point of the new app, dubbed Douyin (“Shaking sound”).

Douyin was a near pixel-level recreation of Musical.ly. Opening the app, you were thrust immediately into a video, with no play or pause buttons. Swipe up, and you could scroll through a seemingly endless feed of 15-second videos, one after another, filling the full screen of the phone’s display. The interface was doubly effective: simple and intuitive enough for anyone to figure out on the first use, while also designed to capture as much data as possible for the recommendation engine. In most vertical scrolls, you’re shown several items at once, making it difficult for the platform to know what you’re looking at or what you think of it. By putting one video in front of you at a time, Douyin could better decipher how the user reacted, and use that data to refine future recommendations.

“TikTok” was a name ripe for virality: It could be pronounced in much the same way across the globe, no matter the language, from Japan to India to Argentina.

Douyin took a canny approach to growth. While many other ByteDance products catered to older or more rural populations, Douyin initially focused on young people from first-tier Chinese cities, the kinds of users whom other people wanted to emulate. “They built influencers,” Erin Huang, one of Douyin’s first contracted creators, told me. “They weren’t shouting, ‘Use our app!’ They were saying, ‘Hey, do you see this person? You can be like them.’” Growth exploded around February 2017, as challenges like “Dance as if taking a bath” went viral on other Chinese social media platforms, drawing in new users and stocking Douyin with more and more content. By May, the app had surpassed one million daily average users. Around the same time, Zhang reached out to the founders of Musical.ly, proposing an acquisition. Zhang believed it would be a perfect fit, according to an investor involved in the deal, because each had what the other needed. ByteDance offered algorithmic capability and business acumen; Musical.ly offered a way onto the phones of millions of American teenagers, the most valuable customers in the world.

According to reporting by Benita Zhang, a prominent Chinese business journalist, Musical.ly put two key conditions on a deal with ByteDance. First, the name of the app had to be changed, to break free from the association with lip-syncing preteens. Second, ByteDance would spend at least $1 billion on marketing. Zhang agreed, and acquired Musical.ly in November 2017 for roughly $1 billion. By that point, ByteDance had already created an international version of Douyin, but had not yet determined how to combine the new acquisition with the existing product.

The company had struggled, too, when figuring out what to name the new app. A team at ByteDance had drawn up a list of English words and whittled down the choices. One possibility was “TikTok.” It sounded cool, but the team reportedly worried that, to Western ears, it would call to mind the 2009 Kesha song. Ultimately the imperatives of ByteDance’s global ambitions overrode any concerns. “TikTok” was a name ripe for virality: It could be pronounced in much the same way across the globe, no matter the language, from Japan to India to Argentina. With a name like that, an app could take over the world.

Image
Credit...Photo illustration by Pablo Delcan

ByteDance’s Beijing offices were located in Haidian, the heart of China’s tech industry, and they were laid out much like any millennial-dominated internet firm. Desks were arranged in long rows, internet-cafe style, with a conference room set in the middle of the floor plan and retro-themed phone booths sprinkled throughout. Every corner had a break room stocked with drinks and snacks. Even the very top executives, like Zhang Yiming, worked out of shared offices.

The fidelity to American business culture ran deep. Zhang was known to quote Jack Welch and Steve Jobs frequently, particularly Jobs’s famous injunction to “stay hungry, stay foolish.” When Zhang organized a book exchange within the company, the title he chose was “The Seven Habits of Highly Effective People.” The first tenet of ByteDance company culture, reinforced through banners and posters throughout the office (including, on occasion, in the bathrooms), is “Always Day 1,” a maxim taken directly from Amazon.

Zhang wore a T-shirt and jeans most days, and insisted that everyone call him by his given name, Yiming — a rarity in the formal and status-obsessed world of Chinese corporations, especially for a high-profile founder. “I hate formality, I hate hypocrisy,” Zhang told an interviewer. He was a frequent Douyin user himself, often creating videos and experimenting with new stickers. At mealtimes, Zhang waited in the same line as the rest of the staff. The idea that executives would have individual elevators, not uncommon at large companies, was “very cheesy,” he once wrote in a note to employees.

But according to interviews with current and former ByteDance employees, who spoke on the condition of anonymity out of concern for professional consequences, the company was caught between the cultures it was trying to bridge. Employees say they were expected to work “996,” meaning 9 a.m. to 9 p.m., six days a week — 72 hours — a standard schedule for Chinese tech companies. During this early period of expansion, calls with overseas offices often ran as late as midnight, and important meetings took place on Sundays. ByteStyle, the company’s code of values, preaches a culture that could have been lifted wholesale from Google or Amazon: diverse, inclusive, radically honest and transparent. But discussing salaries was “a line drawn in blood,” one former employee said, and speaking with the press was absolutely forbidden. The structure was flat, especially by Chinese standards — ByteDance abolished titles for senior positions, and let all employees access other employees’ metrics, including Zhang’s. But it was still clear in which direction orders flowed, and managers were rarely questioned.

“ByteDance is run like a machine,” a former employee said. In China, the company is nicknamed the Super App Factory, in recognition of its streamlined system for pumping out new products. (By one count, ByteDance had more than 140 apps under its umbrella between 2018 and 2020.) The high level of organization and systematization is one of the company’s strengths, because it allows for rapid progress and growth. But it can also be cold and dehumanizing. “Your goals are publicized, and they instill the mantra that your peers are your competitors, not your friends,” the employee said. “It’s like a boiler room, a Wall Street boiler room.”

When the company’s international expansion began, all staff members were told to learn English. Zhang was learning, too, and he sometimes mentioned books he had heard on “Speak English,” a popular E.S.L. app, like the Eckhart Tolle book “The Power of Now.” In 2020, ByteDance hired 40,000 new employees — an average of 150 every working day — many of them outside China, and most under pandemic conditions. Some Chinese employees bristled at the consequences of the expansion abroad. “A lot of Chinese employees may have been working for ByteDance for years, and they didn’t want to start studying English or talking to foreigners or switching the company values,” another former employee told me. “For a lot of people in the Beijing office, they felt they were losing their company to Yiming’s conquest of foreign markets.” Some Chinese employees were reportedly upset at the way that foreign hires described themselves as only working for TikTok in their LinkedIn profiles, with no mention of ByteDance.

The integration was complicated for the foreign employees too — particularly those who came to ByteDance from senior roles at big American tech companies. Having been promised autonomy and independence, they found it could be difficult to accept that ultimate authority rested in Beijing. “America has been so used for so long to being the standard setter and arbiter of business practice, to be the home market and the HQ, that it’s not in the American psyche to be one of the regions,” the second former employee said. “The Americans aren’t used to not having their way.”

For the foreign employees at the Beijing headquarters, the role of cultural translator was an unavoidable part of the job. When ByteDance tried to internationalize one of its short video products, the first former employee recalled, he was called in to consult. In China, the product was known as Xigua Shipin (“Watermelon Video”), and the internationalization team announced that they had chosen an overseas name: “Ripe Melons.” He told them that they couldn’t call it that. “They said, ‘Why?’” the former employee said. “I said, ‘Just trust me, you can’t.’ They thought it was a great name. I said, ‘Melons are a slang word for women’s breasts.’ They’re like, ‘No, it’s melons that are fresh.’” The product was eventually named BuzzVideo.

Gliding across cultures as a kind of internet-era anthropologist was part of what made working at TikTok interesting and novel. When the app was first introduced, every country and every market had a slightly different proclivity. Thai users liked videos of people dancing at school; Japanese users preferred funny videos about otaku, young people obsessed with anime, manga and video games; Vietnamese users especially enjoyed deft camera work. The United States proved harder to crack, until TikTok’s product managers let the users drive the creation of a new category — Americans, it turned out, had an unusual attachment to memes.

But often, ByteDance’s rapid foreign growth resulted in a strange mash-up. “TikTok’s culture is incredibly Chinese in a way contrary to the advertising materials, in a way that’s jarring to foreigners,” the second former employee said. “But on the flip side, it’s a much more foreign tech company than most Chinese people have worked in before.” In Beijing and foreign offices alike, turnover was often high, as employees burned out on the long hours, the coordination across time zones and the juggling of cultures. But success eventually brought its own kind of stability. “It’s become a mainstream tech firm — we’re getting people from Google, Facebook, Snapchat, consulting, blue-chip firms,” a current American employee said. “It no longer feels in any way like a pariah Chinese company.”

Image
Credit...Photo illustration by Pablo Delcan

While ByteDance was pushing abroad, China was changing. When Zhang first started the company, in 2012, Xi Jinping had not yet taken power, and it was still possible to envision the country moving on a path, however gradual and uneven, toward greater reform and openness. But under Xi, that hope has been smothered. Since taking office, he has embarked on a far-reaching reassertion of state power. Among Xi’s priorities has been a widespread crackdown on big tech, one part of a broader effort to rein in private companies. While most big American tech firms had long been barred from operating inside China, the country’s domestic giants had been tolerated, even nurtured. But fearing that tech firms’ wealth and influence could pose a threat to party power and economic stability, the Chinese government has punished, fined and regulated the sector into line with party objectives. One by one, China’s most prominent firms and their founders came under the hammer.

When Ant Group, the financial services arm of Alibaba, was nearing an I.P.O. that was anticipated to be the world’s largest, the government ordered it suspended, and soon after announced a probe into Alibaba’s supposed monopolistic practices. The company’s charismatic founder, Jack Ma, briefly disappeared. In April 2021, Chinese regulators called in the leaders of nearly three dozen of the country’s largest tech companies, ordering them to “learn from Alibaba” and conduct a “comprehensive self-inspection” within a month. In July, Didi, the Chinese version of Uber, was ordered to halt new user registrations just two days after the firm’s American I.P.O.; the company announced that it would delist a few months later. New rules unveiled soon after required internet companies with more than one million users to seek government clearance before listing on overseas stock exchanges. Once seen as the embodiment of China’s economic dynamism and the carrier of Chinese soft power abroad, the country’s tech sector was reminded that it exists at Beijing’s mercy.

A raft of legislation has made clear the government’s priorities. The Cybersecurity Law and National Intelligence Law, each of which took effect in 2017, created affirmative legal responsibilities for Chinese companies and citizens to assist the investigative and intelligence-gathering activities of state organs. “Any organization or citizen shall support, assist and cooperate with state intelligence work in accordance with the law,” the National Intelligence Law states, “and maintain the secrecy of all knowledge of state intelligence work.” In 2021, two new laws on data security asserted the extraterritorial reach of the Chinese state over any data on Chinese citizens anywhere in the world.

‘The regulatory authorities have provided us with much guidance and assistance, but in our hearts we failed to properly understand and recognize it.’

Communist Party cells, a fixture of China’s state-owned enterprises, have taken on a larger and more prominent role throughout the economy. Under Chinese law, all organizations with more than three party members are required to form a party cell, which reports directly to the party bureaucracy and often exerts control over business decisions. Companies frequently publicize the activities of their Party cells as a way of currying favor with the state. “At the end of the day, the Chinese state holds all the cards,” Jordan Schneider, a China analyst at the Rhodium Group and host of the “ChinaTalk” podcast, said. “Firms and their leadership have learned that pushing back too much on government demands can have severe consequences.”

In late 2017, as ByteDance passed $20 billion in valuation, government regulators ordered that updates to several popular verticals on Jinri Toutiao be halted for 24 hours. The app, regulators said, was “spreading pornographic and vulgar information” and “causing a negative impact on public opinion online.” ByteDance later announced that it would hire 2,000 new “content reviewers,” with preference given to party members. The company also shut down the gossipy Society section and created a new vertical called New Era, featuring state media coverage.

A few months later, ByteDance pulled one of its short-video products from app stores after state media accused the platform of glorifying teen pregnancy and hosting advertisements for fake products. Within a week, Jinri Toutiao was temporarily pulled from app stores as well. Regulators then ordered ByteDance to shut down the company’s oldest product, a humor app called Neihan Duanzi (“Subtle Jokes”). The app’s content — a mix of not-so-subtle jokes and sketch videos — had “caused strong dislike among internet users,” the authorities said.

At around 4 a.m. Beijing time the morning after the measures were announced, Zhang posted a long apology. He said he had been unable to sleep, filled with remorse and guilt. Neihan Duanzi had failed to live up to “core socialist values,” Zhang continued. “Over the past few years, the regulatory authorities have provided us with much guidance and assistance, but in our hearts we failed to properly understand and recognize it.” For that, he was sorry — and grateful for the government’s leadership. “As a start-up company developing rapidly in the wake of the 18th National Congress of the Chinese Communist Party, we profoundly understand that our rapid development was an opportunity afforded us by this great era,” Zhang wrote. “I am grateful for this era. I am grateful for the historic opportunity of economic reform and opening; and I am grateful for the support the government has given for the development of the technology industry.”

In order to ensure that ByteDance would improve its understanding and implementation of the “four consciousnesses” of Xi Jinping Thought and hew to “public opinion guidance,” the company would “strengthen the work of party construction,” deepen cooperation with party media and “strengthen the editor-in-chief responsibility system” — a role that, only a few years earlier, Zhang said ByteDance did not need, since it would privilege the preferences of one individual over the aggregated wisdom of all the company’s users.

The contrast with Zhang’s blog post from a decade earlier, when he criticized the expulsion of Google, was stark. It is unlikely that Zhang, who was not a party member, had any great love for the party’s demands. Perhaps he had changed; he now had a multibillion-dollar business to consider, and the livelihoods of tens of thousands of employees. But either way, China certainly had.

Major American platforms at first saw little to worry about in TikTok. No Chinese platform had ever truly captured the American market, and for all that TikTok had gained from the Musical.ly deal in terms of users, it seemed likely to pay for in terms of future growth. Musical.ly had swept up a preteen audience and then stagnated; there was little reason to think TikTok would fare any differently. Besides, TikTok was not really a social network at all. The reason people wanted to be on Facebook, Snap or Instagram was because their friends were on it. “People thought the social network is where the moat comes from; it’s why Facebook is hard to compete with,” an early adviser to ByteDance told me. The major players in American social media were considered so entrenched and immovable that they faced calls to be broken up on antitrust grounds.

If anything, the flush newcomer from China seemed like good news. In 2018, ByteDance spent almost $1 billion in advertising for TikTok, with a budget that reportedly doubled for three straight quarters. TikTok blanketed Facebook, Instagram, Snap and other American social media platforms with ads. TikTok upped the pace again in 2019, reportedly doling out $3 million a day in U.S. advertising alone, a majority of it on Snap. Executives at Snap, who believed their niche as a messaging service was insulated from any threat by TikTok, welcomed the spending. “We were buying so many ads, the money was eye-popping,” a former ByteDance employee said. The response from Snap, he recalled, was, “What can we do to keep this coming?”

At the time, TikTok’s 30-day user-retention rate was rumored to be just 10 percent. By the standards of most social media companies, TikTok was flailing — burning money with nothing to show for it. But ByteDance was playing a different game. Before it could take off, TikTok’s recommendation algorithm needed to be trained to know what was cool — this time for an American audience. The data from Musical.ly’s users provided a first tranche, but in order to make TikTok a compelling product, the algorithm needed to feed on as much data as possible. “It was about getting people, people, people,” a former executive at an American social media company told me. “The more people we get, the more it learns and the more people we can get. That’s the flywheel.”

By mid-2019, the other platforms began to realize that TikTok was not going away. The app had eclipsed 100 million daily average users worldwide, and minted its first bona fide superstar in the artist Lil Nas X, establishing TikTok as a launching pad for musical fame. Then, with the pandemic driving whole countries into lockdown in 2020, TikTok prospered like never before. According to reporting in the Chinese business press, TikTok gained 110 million daily average users between March and April alone. In Iraq, TikTok counted 40 percent of the country’s total mobile-internet population as monthly active users, despite conducting no advertising, promotion or outreach there. The mobile-intelligence firm Apptopia estimated that TikTok was downloaded 89 million times in the United States in 2020, eclipsing even Zoom. In just the first half of the year, according to the research firm Sensor Tower, the app had more than 620 million downloads worldwide.

For Facebook in particular, the threat was acute. The platform was bleeding young users, and engagement was plunging; its fortunes seemed to be cratering just as TikTok’s surged. ByteDance had also been poaching Facebook executives, including Instagram’s head of public policy for the Asia-Pacific region, and Facebook’s public policy leads for Indonesia and Japan.

In a speech at Georgetown University in October 2019, Zuckerberg drew the battle lines. He draped Facebook, which had tried unsuccessfully for years to enter the Chinese market, in the American flag, portraying the contest between the two platforms as one of freedom versus oppression. “China is building its own internet focused on very different values, and is now exporting their vision of the internet to other countries,” he said. He noted the strong encryption protections on WhatsApp, a Facebook product, and its use by protesters around the world. By contrast, he said, TikTok was said to be censoring discussions, even in the United States. “Is that the internet we want?” he asked.

According to reporting by The Wall Street Journal, at a dinner with President Trump during the same Washington visit, Zuckerberg pressed Trump on the threat to American businesses from Chinese internet companies. Stopping Chinese platforms should be a higher priority than concerns over Facebook’s hegemony, he reportedly argued. (A representative for Facebook said that Zuckerberg did not remember discussing TikTok in particular.) Facebook set its substantial lobbying operation to work, meeting with lawmakers and White House staff members to fan the flames against the Chinese app.

TikTok had opened a Washington office earlier that year, with a single employee. Sitting at the intersection of three of the most contentious issues in American politics — China, big tech and social media — TikTok knew it would face scrutiny. “There was no way this wasn’t going to be a tough slog,” a person familiar with TikTok’s lobbying operations said. “It wouldn’t matter who was here or what they did; just the basics of the company make it very hard.”

When the Facebook campaign began, the office was just starting its Washington outreach. According to someone who was present at a September 2019 event at the Langham Hotel in Pasadena, Zhang asked officials from SoftBank, the Japanese conglomerate and one of ByteDance’s biggest investors, for guidance on how to staff up a D.C. operation at a rapid clip. The pace of hiring quickened later in the fall as Zuckerberg pressed the case against Chinese companies. The resulting ByteDance lobbying team was a who’s who of Washington, composed of former congressional aides to leaders in both parties, including Speaker of the House Nancy Pelosi; Kevin McCarthy, House minority leader; Chuck Schumer, Senate majority leader; and Jim Clyburn, House Democratic whip. From just over $500,000 in 2019, ByteDance’s lobbying budget grew nearly eightfold in 2020, then almost doubled again the next year. “They literally went from one person to, like, 30 in a year and a half,” the person familiar with TikTok’s lobbying operations told me. “It was one of the biggest ramp-ups I’ve ever seen.”

In mid-July 2020, President Trump, Vice President Pence and members of the cabinet gathered in the Cabinet Room of the White House for a meeting. Among the agenda items was TikTok. After a briefing by the deputy national security adviser, Matt Pottinger, there was a short discussion. “It was clear that there was more or less unanimity that TikTok is a national security threat,” a person present said. “The discussion was how to deal with it.”

Administration officials had first begun outreach to Capitol Hill about TikTok in early 2019. At the time, the administration was preparing a series of actions against Huawei, the Chinese telecommunications firm, and TikTok seemed to evoke some of the same concerns about data privacy and Chinese government influence. But while TikTok was growing in popularity, it was still confined to a small enough audience to keep it from being a government priority. “There was hope that it would be a fad, like Myspace,” said a Trump White House official, who requested anonymity to discuss internal governmental deliberations. “That it would just fade out.”

By 2020, it was clear that wouldn’t happen. TikTok’s growing dominance provoked two primary concerns: First, that the data gathered on American users could be accessed by Beijing and deployed for the purposes of blackmail, harassment or espionage. Second, that the algorithm itself could be used to advance the Chinese government’s foreign-policy goals, whether by promoting content favorable to Beijing or by suppressing views deemed objectionable. The company had already been accused of censoring content considered politically sensitive in China, as well as removing or burying videos related to Black Lives Matter, protests in Hong Kong and the repression of the Uyghurs in China’s Xinjiang Province. (TikTok has said that these were temporary issues that do not reflect the current workings of the app.)

Trump’s cabinet considered three options. The first was to let the Committee on Foreign Investment in the United States take the lead. CFIUS — an interagency panel tasked with assessing the national-security implications of investment from overseas — had already opened an investigation into ByteDance’s acquisition of Musical.ly. The committee would make a recommendation to the president, suggesting either a full divestment of Musical.ly — a death sentence for TikTok in its most important market — or mitigation measures that could allay the government’s concerns. The second option, favored by the U.S. Treasury secretary, Steven Mnuchin, was to allow TikTok to remain a Chinese company but operate in partnership with an American company, which would host its data servers on U.S. soil.

The third option was an outright ban of the app. The government of India had already banned TikTok and dozens of other Chinese apps on national security grounds, following deadly border clashes with China. In the United States, a similar move would require a strong and well-developed legal theory, taking into account First Amendment concerns and the distinction between objectionable publishers, which cannot be banned, and a foreign-owned platform.

Presented with three choices, Trump chose the flashiest and most legally dubious: an outright ban. The previous month, TikTok users pranked Trump’s re-election campaign by organizing people around the world to register for a rally in Tulsa with no intention of showing. Brad Parscale, the chairman of Trump’s re-election campaign, tweeted that they had received more than one million ticket requests; about 6,200 people ultimately showed up. The botched rally seemed to have put TikTok on the president’s radar. “That was a pretty major factor in the executive order,” said a former Trump administration official who requested anonymity to speak candidly about the White House’s decision-making. Trump signed an executive order preventing further downloads and updates of TikTok in the United States if the company was not sold to an American buyer within 45 days. Eight days later, CFIUS concluded its investigation, recommending that the president order a divestment.

The weeks following Trump’s announced ban were a circus to rival anything the administration had produced to date. Seeing a chance to snatch the world’s most popular app at a fire-sale price, suitors swooped in. Seemingly every major American tech company was rumored to be pursuing a deal, including Microsoft, Apple and Alphabet, the parent company of Google. Eventually, the contours of an agreement were reached. Walmart and Oracle would jointly purchase a stake in a new U.S.-based entity, with ByteDance remaining the majority shareholder. Oracle would oversee the app’s data, ensuring that the personal information of American users was stored only in the United States. At the same time, ByteDance reportedly redoubled efforts already underway to move more of the company leadership to Singapore.

Zhang reportedly learned of the proposed ban when a friend sent him a link to an interview with Trump. From his base in Beijing, Zhang kept American hours, huddling with advisers and investors about how to respond. He wrote a letter to his employees, reiterating his desire to lead “a trustworthy global company.” In China, Zhang’s letter received hundreds of sharp replies, condemning him as a traitor, a coward and an American stooge, sometimes referencing his youthful blog posts that favorably described American culture and politics. Zhang hardly deserved to be called Chinese at all, some angry online commenters said — really he was a jingmei fenzi, an “American at heart.”

And then, just as capriciously as it had burst forth, the drama subsided. After the 2020 election, the proposed deal was shelved, and the Trump administration seemed to forget about the issue entirely. Though Biden quickly rescinded Trump’s executive order, the CFIUS divestment order remained in place. TikTok was left in limbo: operating normally for all appearances, but with the lingering threat of a ban or other government action constantly overhead.

The Biden administration came into office promising to repudiate its predecessor in both style and substance. But China policy — and China tech policy in particular — has represented a clear continuity. The Biden administration has come to see tech advancement as a zero-sum game that the United States must not lose. The centerpiece of Biden’s China tech policy thus far has been a set of sweeping restrictions on the sale of the most advanced semiconductor chips and chip-making equipment to China. Semiconductors are the lifeblood of the digital age, powering every industry and field in which the United States and China are competing, from autonomous cars to artificial intelligence, supercomputing and advanced missiles. The new rules were striking for their unprecedented breadth, cutting off not just the chips themselves but also access to any U.S.-made software or components needed to make the chips.

Under the policy, any “U.S. person” — citizens, permanent residents, anyone who lives in the country and U.S. companies — must receive a government license to work with a Chinese company in the field of advanced-semiconductor production. Taken together, the restrictions seem designed to smother China’s tech industry before it can catch up to America’s. Tech decoupling between the United States and China, once seen as extreme option, is now debated only in its details: when, how and where. “This is going to manifest across trade, investments, people and ideas,” said Schneider, the Rhodium Group analyst. “You’ll likely end up seeing decoupling as a slow and steady march on every dimension.” But in the middle of the Biden administration’s China tech policy sits a TikTok-size hole.

If TikTok has escaped the scrutiny faced by other Chinese companies (or even other American social media giants), it is in part because the user base skews so young. According to the Pew Research Center, two-thirds of 13-to-17-year-olds in the United States use TikTok. Few lawmakers or regulators even understand TikTok. The app’s opacity has also offered a shield. Unlike Facebook and Twitter, TikTok does not share data with researchers or allow outsiders to study the platform. (Last month, TikTok said it was preparing a beta version of a platform-research tool.)

The controversies swirling around TikTok have done little to blunt the app’s ascent. In September 2021, the company announced that it had reached one billon active monthly users, meaning that it had been adding new users at an average rate of nearly 550,000 per day for five straight years. Even with the ban in India, TikTok was the most-downloaded and highest-earning nongame app in the world for the first half of this year. Between Douyin and TikTok, more than one billion people around the world use a ByteDance app every day. On Douyin alone, the average cumulative use time each day accounts for something like 90,000 years.

‘If you look at the people who draw the analogies between Google and Facebook and TikTok, they’re either unsophisticated or they have an ax to grind in favor of TikTok.’

Facebook, which renamed itself Meta in 2021, is still pressing Washington for help. According to emails viewed by The Washington Post, the company has hired one of the biggest Republican consulting firms in the country to lead a nationwide public relations campaign against TikTok. The firm, Targeted Victory, has placed opinion columns and letters to the editor in regional newspapers, encouraged journalists and politicians to dig into TikTok and helped spread damaging news stories. The overall aim is to “get the message out that while Meta is the current punching bag, TikTok is the real threat, especially as a foreign-owned app that is #1 in sharing data that young teens are using,” a director for the firm wrote in a February email.

TikTok has proved a ripe target. Over the summer, BuzzFeed reported on leaked audio from dozens of internal company meetings revealing that, contrary to TikTok’s public assertions, data on American users was still routinely accessed by China-based employees. Soon after, BuzzFeed reported that ByteDance had used TopBuzz, a now-shuttered American news app modeled on Toutiao, to push pro-China content to users, while also censoring stories critical of Beijing. (ByteDance has denied this, calling it “ridiculous.”) More recently, Forbes found that Chinese state media accounts were flourishing on TikTok, often by promoting attacks on specific U.S. politicians and the state of American institutions in general. Forbes also reported that a team at ByteDance headquarters planned to use TikTok to track the location of specific American users — exactly the nightmare scenario that critics had warned about. Taken together, these stories have only amplified concerns that TikTok cannot be trusted with its power over American data and attention spans.

The company’s response has done little to allay concerns. Instead of acknowledging the unique hazards of their situation, company officials have sought to distance TikTok as much as possible from its Chinese origins and ownership. TikTok has claimed, for instance, that it is not a Chinese company because the legal entity that owns TikTok and all of the businesses in China is actually registered in the Cayman Islands. TikTok and ByteDance executives have also repeatedly said, including in sworn testimony to Congress, that TikTok has never provided user data to the Chinese government, nor would it if asked — an assertion that the company would knowingly violate Chinese law. “If you look at the people who draw the analogies between Google and Facebook and TikTok, they’re either unsophisticated or they have an ax to grind in favor of TikTok,” said Dan Harris, a lawyer who works with foreign companies in China and writes the China Law Blog. “Most serious people see a difference. It doesn’t mean they’re all great or all bad, but there is a difference.”

TikTok is reported to be making progress on a deal with the Biden administration that would allow the app to retain its Chinese ownership, but keep its American user data on servers in the United States. That arrangement seems unlikely to satisfy anyone, but all the available solutions are imperfect. An outright ban, especially one targeting Chinese companies writ large, risks looking like Sinophobia, and also — somewhat counterintuitively — like acquiescence to the Chinese Communist Party’s view that every citizen and entity in China is a willing appendage of the party. Yet to turn a blind eye to the potential risks posed by a company like TikTok is to ignore the political, economic and social infrastructure of control that the Chinese government under Xi has spent more than a decade constructing.

Whatever happens with ByteDance, the lessons for the next Chinese entrepreneur are sobering. “He can do what ByteDance is trying to do: get a Singaporean passport and incorporate there,” said Ivan Kanapathy, a former director for China, Taiwan and Mongolia on the National Security Council staff. “There’s no answer if he’s in China. If he wants to be a global tech company, that’s it. You can’t have both. If you want the China market, go to China. If you want the West, go West. That’s where we’re going. I have no doubt.”

In May 2021, Zhang Yiming announced that he would transition out of his role as C.E.O. of ByteDance over the next six months. It was time for him to take on a new role, Zhang said in a letter to employees, both to avoid the trap of becoming too central to the organization, and to help spur the type of longer-term thinking that would lead to ByteDance’s next breakthrough. “The truth is, I lack some of the skills that make an ideal manager,” he wrote. He preferred to spend time alone, reading, being online and daydreaming about the future.

A few months later, news broke that the Chinese government had taken an ownership stake in a ByteDance subsidiary. The subsidiary held operating licenses for some of the company’s most important Chinese businesses. Though the size of the stake was small — just 1 percent, divided between the China Internet Investment Fund; China Media Group, controlled by the Communist Party’s propaganda department; and the Beijing municipal government’s investment arm — the implications were unavoidable. The Chinese government took one of three seats on the subsidiary’s board, wielding a level of influence incommensurate with its nominal stake.

Zhang is reported to have spent most of his year in Singapore, ostensibly to escape the onerous restrictions and lockdowns of China’s “zero Covid” policy. But it’s not hard to see the relocation as driven by other factors. “I don’t think Zhang Yiming is all that interested in being a tool of any government,” Schneider said. “I think he wants to build an online empire and make a lot of money. Most entrepreneurs in most corners of the globe don’t have to worry about their success being branded a tool of the state.”

In 2016, just after the release of Douyin, Zhang gave a long and wide-ranging interview to the Chinese magazine Caijing. Though Toutiao was a well-established success, Zhang was not yet a household name in China, and he was almost totally unknown overseas. Over the course of the interview, Zhang talked about human nature, corruption, integrity, books that had influenced him, the importance of delayed gratification and much else. It was his dream, Zhang said, to “learn English well,” though running a company left little time for anything besides work.

Toward the end, the interviewer asked Zhang about his reputation for being extremely realistic, “like a robot.” Zhang replied that failing to face reality always created problems for people. “The best way to predict the future is to create it,” he said, “but only if you face it.”


Alex W. Palmer is a contributing writer for the magazine. His 2019 cover story on the D.E.A.’s investigation of a North Dakota teenager’s death by fentanyl overdose was a finalist for a National Magazine Award. Pablo Delcan is a designer and art director from Spain. In 2014 he founded Delcan & Co., a design studio based in New York.

A version of this article appears in print on  , Page 22 of the Sunday Magazine with the headline: How TikTok Became a Diplomatic Crisis. Order Reprints | Today’s Paper | Subscribe

Advertisement

SKIP ADVERTISEMENT