Examining TikTok
Its problems may only just be starting
TikTok is simply too consequential to ignore—as a recent spate of unflattering media coverage makes abundantly clear. Because the platform has proven freakishly adept at a) building up a monstrous global user base, b) expertly monopolizing and monetizing attention, and c) harvesting ungodly amounts of data to further fine-tune its powerful recommendation algorithm, TikTok is both one of the most consequential actors in the tech space today and a lightning rod for public criticism.
I find TikTok, and the heated debate swirling around it, particularly fascinating because it hits on many of the core themes explored in this newsletter: cross-border data flows, polarization and misinformation, AI, and the battle between the United States and China for technological supremacy. For better or worse, TikTok embodies the 2022 zeitgeist. Its conquests and challenges will illuminate the technological and geopolitical order we inhabit going forward. I think the forecast is grim.
The obvious starting point here, as Ezra Klein wrote earlier this summer, is the TikTok-China dynamic:
TikTok is owned by ByteDance, a Chinese company. And Chinese companies are vulnerable to the whims and the will of the Chinese government. There is no possible ambiguity on this point: The Chinese Communist Party spent much of the last year cracking down on its tech sector. They made a particular example out of Jack Ma, the high-flying founder of Alibaba. The message was unmistakable: Chief executives will act in accordance with party wishes or see their lives upended and their companies dismembered.
With that context in mind, it feels like we’re fast approaching a breaking point. On the one hand, tens of millions of Americans’ (especially Gen Z) lived experiences and perceptions of reality are intimately mediated through TikTok interactions. On the other, US-China relations have nosedived over the past decade. In my view, the relationship will worsen further in the years ahead. As I’ve noted before, technology and information security are central fault lines in this geostrategic rivalry. Machine learning, data access, and technological interdependence—all touched on by TikTok’s enormous footprint—present potential points of conflict and areas of concern.
To be clear, the toxifying geo-technological dynamic around TikTok is already in motion. Let’s remember that two years ago the Trump administration unsuccessfully attempted to either force TikTok’s sale or ban it outright. In fact, the US armed forces officially banned the app from government devices around early 2020. More recently, President Biden honed in on the People’s Republic (with an eye toward TikTok) when he issued an updated executive order earlier this summer addressing the “extraordinary threat to the national security, foreign policy, and economy of the United States” posed by “connected software applications that are designed, developed, manufactured, or supplied by persons owned or controlled by, or subject to the jurisdiction or direction of, a foreign adversary.”
We’ve already seen this story play out between Beijing and New Delhi. Last summer, Indian officials banned TikTok along with 58 other Chinese apps in a move widely interpreted as retaliation for the violent 2020 Galawan Valley border clash. As I wrote in India & The Tech War a few weeks back:
I find it unsurprising that Chinese telecom equipment providers like ZTE and Huawei and platforms like TikTok and WeChat are feeling the effects of the Sino-India breakdown most acutely. New Delhi’s crackdown reflects the same fears held by American, Chinese, Japanese, and European officials over the risks of digital interdependence.
Whether or not those fears are fully realized—in the form of shutting down a network, tweaking an algorithm, or weaponizing data flows—is beside the point. The fact that these things could happen, that “they” could do it to “us,” is shifting national priorities from building a “borderless and open” digital commons to better safeguarding domestic information security—even if this newfound emphasis on control comes at the expense of efficiency, cost, or stifled innovation potential.
ByteDance’s potency
I want to set aside the safety and security concerns for a moment and draw attention to the fact that what (TikTok’s Chinese parent company) ByteDance has accomplished is genuinely impressive from a product development standpoint. The always astute analyst Ben Thompson over at Stratechery lays out this “secret sauce” as such:
ByteDance’s breakthrough product was a news app called TouTiao; whereas Facebook evolved from being primarily a social network to an algorithmic feed, TouTiao was about the feed and the algorithm from the beginning. The first time a user opened TouTiao, the news might be rather generic, but every scroll, every linger over a story, every click, was fed into a feedback loop that refined what it was the user saw.
Meanwhile all of that data fed back into TouTiao’s larger machine learning processes, which effectively ran billions of A/B tests a day on content of all types, cross-referenced against all of the user data it could collect. Soon the app was indispensable to its users, able to anticipate the news they cared about with nary a friend recommendation in sight. That was definitely more of a feature than a bug in China, where any information service was subject to not just overt government censorship, but also an expectation of self-censorship; all the better to control everything that end users saw, without the messiness of users explicitly recommending content themselves (although that didn’t prevent ByteDance CEO Zhang Yiming from having to give a groveling apology for giving users too much low-brow content).
ByteDance’s 2016 launch of Douyin — the Chinese version of TikTok — revealed another, even more important benefit to relying purely on the algorithm: by expanding the library of available video from those made by your network to any video made by anyone on the service, Douyin/TikTok leverages the sheer scale of user-generated content to generate far more compelling content than professionals could ever generate, and relies on its algorithms to ensure that users are only seeing the cream of the crop.
So here’s what happened in a nutshell: TikTok’s unmatched algorithmic excellence plus its reliance on a platform-wide “content graph” over the traditional, network-driven “social graph” to surface new videos—matched with the prioritization of short-form video over text and an easy-to-use, intuitive interface for content creators—allowed it to rapidly upend America’s social media ecosystem and set the competition scrambling.
From The Economist:
Beneath TikTok’s simple interface lies fearsomely advanced artificial intelligence (ai). Its knack for learning what people like helped TikTok sign up its first 1bn users in half the time it took Facebook. In America the average user spends 50% longer on the app each day than the typical user spends on Instagram. TikTok’s revenues are expected to reach $12bn this year and $23bn in 2024, drawing level with YouTube’s. Young creators are flocking to the app—along with some older ones.
The effect on competition has been dramatic. In 2020 American trustbusters sued Facebook, now known as Meta, for its alleged dominance of social media. Today such worries look eccentric; Meta has been particularly hard-hit as tech stocks have taken a beating, and the firm is re-engineering its products to mimic TikTok. America often accuses China of copycat capitalism. Now the boot is on the other foot.
Echoing that sentiment is Kai-Fu Lee, arguably the leading evangelist for China’s mobile internet and AI ecosystem, who bluntly puts it: “if nothing else, TikTok crushes the outdated theory that Chinese companies are merely copycats.”
I think this is a crucial point. While most Americans are undoubtedly aware of their deep entanglement with the People’s Republic—undergirded by the heft of Chinese labor, capital, consumer market size, and critical role in global supply chains—there has generally been a lack of identification with, much less fondness for, Chinese brands, businesses, or technology platforms. Korea has Samsung, Hyundai, and K-pop. Japan has Sony, Toyota, Mitsubishi, and Pokémon. Until TikTok’s arrival, China, by contrast, lacked anything with similar resonance. “National champion” telecom gear provider Huawei, perhaps the one notable exception, primarily operates in the B2B space and is something of a global pariah.
While it’s long been taken as axiomatic among informed outside observers that “China cannot innovate,” much less develop successful, truly beloved technology products for a foreign audience, TikTok has proven every bit the worthy rival of its Silicon Valley peers; many of whom are moving to emulate the platform’s content graph surfacing model.
TikTok doesn’t read to many observers as the feel-good success story of a plucky upstart disrupting the social media ecosystem, empowering creators, and potentially revolutionizing the US retail sector. Success has bred worry—often bordering on panic—and fueled calls for the app’s removal from the Google and Android stores. Nor are those concerns strictly limited to the halls of power in DC. At the recently concluded Code Conference—the annual LA confab for tech luminaries—Steven Levy over at Wired noted that “the invocations of TikTok were not that of a haunting, but of a looming terror.”
China's (and tech’s) irresolvable trust dilemma
I’ll start with what I think is an important point: in an environment defined by high levels of tribalism, polarization, and zero-sum thinking (basically, the world in 2022), capabilities (means) matter just as much as intentions because “they” will always be assumed to be acting in bad faith. This is doubly true in the domain of international relations, where intense cultural and political differences, the oft-competitive nature of the nation-state system, and the prioritization of national security create incentives to conceal intentionality and throw up massive barriers to building trust. In a trust-deficient environment, the lack of a smoking gun won’t convince many people that it doesn’t exist or won’t exist in the near future.
I will go further and say this is triply true when it comes to information and communications technology. It’s powerful, complex, interconnected, opaque, fast-moving, often intangible, and essential to modern life. ICT’s omnipresence raises the specter of unceasing digital surveillance and an accompanying loss of agency and autonomy. These technologies offer up seemingly endless points of potential leverage, manipulation, and vulnerability; all lending themselves to a pervasive, irresolvable sense of paranoia. The attack surface—defined by Crowdstrike as “the collective of all potential vulnerabilities (known and unknown) and controls across all hardware, software and network components”—is simply massive and virtually impossible to offset. To my mind, this is why ICT is at the epicenter of 21st-century political, social, and geostrategic conflict.
All of this circles back to ByteDance/TikTok. We’re talking about a prominent Chinese company with a massive US footprint operating at a moment in time when Sino-American ties are at their lowest ebb in decades. When the two superpowers look at one another, more often than not, they are ascribing the most hostile possible intentionality to one another’s current and future courses of action.
In such an environment, the prevailing frame around commercial and technological entities operating across national borders shifts: once (mostly) seen as conduits for mutually beneficial engagement, they can quickly become recast as malicious agents acting on behalf of their home governments. As the NYT wrote recently, President Xi is “increasingly obsessed with tech ‘choke points’ that have left China vulnerable amid the trade war with the United States, including bans on companies like ZTE, the potential war over the Taiwan Strait and Russia’s invasion of Ukraine.”
Of course, mitigating against the risks of “weaponized interdependence” has long animated China’s political decision-making. China famously expends massive amounts of manpower and resources maintaining its Great Firewall, blocks out US social networks like Facebook and Twitter, hives off strategically sensitive areas like petrochemicals, electricity generation, and segments of the ICT sector from foreign competition, and plows massive amounts of financial firepower into building up viable national alternatives. Edward Snowden’s revelations back in 2013 reinforced long-standing fears over pervasive American surveillance and turbocharged Beijing’s drive for greater levels of information control and techno-sovereignty.
What’s been equally striking, and this now appears true across both the Trump and Biden administrations, is Washington’s full-throated embrace of techno-nationalism—via the use of industrial policy tools (CHIPS Act), trade barriers, and export controls—to both accelerate Sino-American decoupling and frustrate Beijing’s technological ambitions.
In the case of TikTok, it faces a double negative: not only is it a very successful and popular product originating from a country seen by many Americans as an implacable adversary…it is a very successful and popular product originating from a country seen by many Americans as an implacable adversary that operates at the bleeding-edge of big data, social media, mobile internet, and machine learning; technologies that will determine the global balance of power in the decades ahead.
Data and privacy
Of course, this isn’t simply a case of unsubstantiated paranoia on the part of a few geriatric China hawks. Reporting on TikTok has raised significant red flags—specifically around data protection.
To be clear, the company has long held that American user data is never stored in China but in the United States with backup redundancy in Singapore. Via “Project Texas,” negotiated with Oracle and CFIUS, “100% of US user traffic is being routed to Oracle Cloud Infrastructure” and the company is “working closely with Oracle to develop data management protocols that Oracle will audit and manage to give users even more peace of mind” while also creating a new, US-led department tasked with sole management over Americans user data.
According to Chris Stokel-Walker, author of TikTok Boom, there is, as of yet, no evidence that US data is being leaked to Chinese authorities. TikTok insists Chinese officials have not requested US user data and that it would refuse any such requests if they were made. Analysis from the University of Toronto Citizen Lab indicates that “TikTok and Douyin do not appear to exhibit overtly malicious behavior similar to those exhibited by malware.”
While TikTok is obviously keen to push out the narrative that it’s proactively safeguarding data and has successfully firewalled its US and China operations, that story erodes a bit under closer scrutiny:
We know that TikTok hoovers up all sorts of information—usage, device info, location, metadata—which can, in fact, be sent to ByteDance HQ. As made clear in its privacy policy, the platform “may share all of the information we collect with a parent, subsidiary, or other affiliate of our corporate group.”
We know, as per Buzzfeed’s bombshell reporting, that “according to leaked audio from more than 80 internal TikTok meetings, China-based employees of ByteDance have repeatedly accessed nonpublic data about US TikTok users” and that access happened “far more frequently and recently than previously reported.” One member of TikTok’s Trust and Safety department baldly stated that “everything is seen in China.” The physical location of user data is ultimately irrelevant if China-based admins retain the ability to access and transfer it at will.
We know that story builds off of similar reporting from CNBC last year which found that “ByteDance has access to TikTok’s American user data and is closely involved in the Los Angeles company’s decision-making and product development.”
Just as important is the macro picture: China’s ruling Communist Party sees itself sees locked in an intense, zero-sum struggle—one composed of geostrategic, technological, and ideological dimensions—with the United States and its allies. As academics Christopher Carothers and Huan Gao point out, to both preserve domestic stability and mitigate against mounting external threats, President Xi has pursued a “security-first” approach to governing China which de-emphasizes economic growth and efficiency in favor of “a vision for maintaining stability” which “calls for extreme political and ideological control.”
Towards that end, China has purposefully built up a robust legal regime—through the 2021 Data Security law and the 2017 National Intelligence and Cybersecurity laws, among other key pieces of legislation—that demands Chinese internet companies like Bytedance “support, assist and cooperate with the state intelligence work” on matters deemed relevant to Beijing’s (broadly defined) national security interests. As the always-excellent China Law Blog puts it:
Under the guidance of the Chinese Communist Party (CCP), the Chinese government is working to create a cybersecurity system with Chinese characteristics. This system is designed to make all networked information that crosses the Chinese border a) transparent to the Chinese government and b) closed to unauthorized access by foreign and domestic hackers and governments not affiliated with the CCP.
The primary goal is to use this system for surveillance and control. Surveillance means acquisition of information. As a result, the transparency of the system means all information that crosses the Chinese border should be available to the CCP and its agents. There are no secrets from the Party. As a result, from the standpoint of an individual or a business entity, whether domestic or foreign, this system is a cyber-insecurity system. As the system is implemented with progressively greater refinement and scope, there will be no place to hide from the eyes of the party.
All of this suggests that Chinese authorities possess the legal and coercive power to compel companies like ByteDance to fork over any and all information relevant to the security of the regime. Given the CCP’s overarching obsession with accessing, protecting, and controlling information, it likely possesses a strong motivation to do so. Even if no conclusive evidence yet exists that this is occurring, the lack of a smoking gun is unlikely to assuage the concerns of US regulators.
Censorship, propaganda, disinformation, and algorithms
Chinese authorities see ByteDance as essential to the country’s technological competitiveness and national security. According to Reuters, Beijing opposed the forced sale of TikTok back in 2020, preferring to see the app shuttered altogether rather than capitulate to US political pressure or risk having details of its all-important recommendation engine shared with American officials or competitors. Earlier this year, Chinese regulators issued updated guidance on the use of recommendation algorithms—the secret sauce integral to the success of many Chinese internet companies—to bring them closer in line with Beijing’s social and national security prerogatives. Taking it a step further, Alibaba, Tencent, Baidu, and ByteDance (among others) were forced to share details of their prized algorithms with the extremely powerful Cyberspace Administration of China (CAC) last month.
All of this underscores the extent to which China cares, and cares deeply, about accessing, controlling, restricting, and manipulating information and ideas to suit its political needs. It’s the reason the party maintains the Great Firewall, blocks out American rivals, and heavily censors content across its domestic social media platforms. ByteDance (and TikTok) cannot help but closely intersect with those political needs given their immense power over data and information. Bear in mind that China forcibly shuttered ByteDance’s massively popular meme and joke-sharing app, Neihan Duanzi, back in 2018 for “vulgar” content that allegedly offended the sensibilities of Chinese internet users.
Of course, one obvious (and problematic) manifestation of China’s influence over its leading internet players is censorship: leaked internal documents reported by The Guardian show that TikTok “instructs its moderators to censor videos that mention Tiananmen Square, Tibetan independence, or the banned religious group Falun Gong” and the platform has faced similar criticism for both censoring and shadowbanning content related to Xinjiang and the Hong Kong protest movement. Even if such moves are not at the direct behest of state officials, they reflect the country’s chilled political environment: over the past few years Chinese authorities have brought its tech sector to heel, self-censorship runs rampant, and powerful players like ByteDance cannot risk falling afoul of the party line.
While data espionage, censorship, and political interference all pose extremely formidable threats to TikTok’s brand identity and future viability in Western markets, to my mind the platform’s biggest struggles going forward will have less to do with what the Chinese government sees (or what TikTok’s users don’t see), but rather what its users do see and how they react. TikTok is a juggernaut battling not just for market share but also “mindshare”—the power to define and mediate reality—in the hyper-competitive attention economy. It is a major source of news and information at a moment in time when social media is at the forefront of the culture war. In the span of just a few years, once-beloved peer competitor platforms like Facebook and Twitter have seen their public standing collapse under withering scrutiny from across the political spectrum over privacy, data security, mental health, censorship, misinformation, and hate speech; the very issues forming the crux of the TikTok backlash.
More existentially is the issue of social media’s effect on democracy itself. Healthy liberal democracies cannot function effectively under conditions of intense political polarization, collapsing trust in social institutions and between fellow citizens, and the lack of a shared baseline in terms of what constitutes objective reality. As scholar Jonathan Haidt wrote in a much-discussed piece for The Atlantic this spring:
It’s not just the waste of time and scarce attention that matters; it’s the continual chipping-away of trust. An autocracy can deploy propaganda or use fear to motivate the behaviors it desires, but a democracy depends on widely internalized acceptance of the legitimacy of rules, norms, and institutions. Blind and irrevocable trust in any particular individual or organization is never warranted. But when citizens lose trust in elected leaders, health authorities, the courts, the police, universities, and the integrity of elections, then every decision becomes contested; every election becomes a life-and-death struggle to save the country from the other side. The most recent Edelman Trust Barometer (an international measure of citizens’ trust in government, business, media, and nongovernmental organizations) showed stable and competent autocracies (China and the United Arab Emirates) at the top of the list, while contentious democracies such as the United States, the United Kingdom, Spain, and South Korea scored near the bottom (albeit above Russia).
Recent academic studies suggest that social media is indeed corrosive to trust in governments, news media, and people and institutions in general. A working paper that offers the most comprehensive review of the research, led by the social scientists Philipp Lorenz-Spreen and Lisa Oswald, concludes that “the large majority of reported associations between digital media use and trust appear to be detrimental for democracy.” The literature is complex—some studies show benefits, particularly in less developed democracies—but the review found that, on balance, social media amplifies political polarization; foments populism, especially right-wing populism; and is associated with the spread of misinformation.
TikTok, by its very nature, plays into all of social media’s worst pathologies as a trust eroder and polarization accelerator. If democracy requires high-quality information and a contemplative citizenry, then TikTok’s format—super short-form video over long-form written text—seems uniquely ill-suited to deliver it.
Layered on top of that is the fact that the platform is awash in coordinated propaganda campaigns, misinformation, and groundless conspiracy theories. These tendencies skyrocket in the run-up to contentious elections and the company’s track record in managing the misinformation problem inspires little confidence. The New York Times notes that “TikTok is shaping up to be a primary incubator of baseless and misleading information” in the lead-up to US midterm elections this fall. As Taylor Lorenz’s recent profile on “Libs of TikTok” makes clear, the brand of hyper-inflammatory content that drives engagement on the platform can quickly extend its reach into the broader information ecosystem.
This is where China and the trust angle come back into play. All of these concerns (censorship, disinformation, polarization, democratic erosion) are not unique to TikTok, but TikTok is unique in multiple ways that make it particularly ominous. Its short-form video format fosters the opposite of an informed, deliberative democratic citizenry at a moment when many analysts fear Western-style liberal democracy is in terminal decline. What users see is dictated by the platform’s opaque recommendation algorithm rather than through your social network, affording TikTok greater latitude for potential manipulation versus rival platforms using the social graph model. That opacity offers up all sorts of avenues for hard-to-detect political interference. From Ezra Klein again:
How comfortable are we with not knowing whether the Chinese Communist Party decided to weigh in on how the algorithm treats these videos? How comfortable will we be with a similar situation in five years, when TikTok is even more entrenched in the lives of Americans, and the company has freedom it may not feel today to operate as it pleases?
Imagine a world in which the United States has a contested presidential election, as it did in 2020 (to say nothing of 2000). If one candidate was friendlier to Chinese interests, might the Chinese Communist Party insist that ByteDance give a nudge to content favoring that candidate? Or if they wanted to weaken America rather than shape the outcome, maybe TikTok begins serving up more and more videos with election conspiracies, sowing chaos at a moment when the country is near fracture.
This is the root of the problem. Barring a miracle, America’s political culture will become more violent, polarized, and dysfunctional in the years ahead—harming both its international power projection capacity and the allure of its democratic example to the rest of the world. Will US officials, for many of whom Sino-American competition is framed as an overt “values conflict” between freedom and autocracy with globe-spanning ramifications, sit idly by as TikTok—already suspected of being under the influence of a hostile ideological rival—spews out content that weakens the country’s democratic fabric and bolsters China’s hand in the battle of ideas?
Balance of power
This brings me to the heated technology competition between the US and China. TikTok, along with Alibaba, Tencent, and a handful of other prominent tech players, represent a global internet that is both more balkanized and “Sinicized” than the US-led model that dominated the post-Cold War period. Their success heralds an emerging bipolarity; an age in which the US no longer monopolizes the industries of the future—specifically machine learning and artificial intelligence—and the rules and norms governing information technology are increasingly set in Beijing.
More viscerally, it raises the specter of a future in which an authoritarian China, by virtue of its technological prowess, supplants America as the global hegemon and can freely impose both its will and values onto the international arena with little fear of US pushback. While I am not at all convinced this is the future we’re destined for, as China’s ongoing innovation problems and ample inbox of domestic and international challenges can testify, the mere prospect that this could be a potential future has engendered considerable fear and pushback.
America’s political class—across the Democrat/Republican divide—has made abundantly clear this is an unacceptable outcome. US Commerce Secretary Gina Raimondo frankly stated that “we can’t let China write the rules around AI.” The National Security Commission report on Artificial Intelligence claims that “China’s domestic use of AI is a chilling precedent for anyone around the world who cherishes individual liberty. Its employment of AI as a tool of repression and surveillance—at home and, increasingly, abroad—is a powerful counterpoint to how we believe AI should be used” and that “China’s plans, resources, and progress should concern all Americans.” The report frames artificial intelligence primarily through the prism of Sino-American strategic competition and makes clear the US and its allies “must win” the AI race in order to preserve both our national security interests and democratic values.
All of this is to say the deck is stacked against TikTok ever making a credible case it doesn’t pose an acute national security risk. Now, to be clear, I don’t see an outright ban on the table for the time being. Shuttering TikTok would carry massive downside political risk for a Biden administration struggling to retain support with millennial and Gen Z voters for whom the app is genuinely popular.
However, as long as it remains tethered to ByteDance, and, by extension, China, I don’t believe TikTok can shake the perception that it’s operating as a cat’s paw for a hostile foreign power hellbent on exploiting sensitive user data, polarizing and undermining American democracy, and helping China gain the upper hand in the foundational technologies of the 21st century. Extrapolating forwards TikTok’s position will likely become ever more untenable in the years ahead.