The “world’s most advanced surveillance technologies,” explains political scientist Minxin Pei, are to be found today in China. These domestic spying initiatives boast powerful data-collection capabilities, as well as ostentatious, science fiction–inflected names: Skynet, Sharp Eyes, Golden Shield. What’s important to Pei’s The Sentinel State: Surveillance and the Survival of Dictatorship in China, though, is that none of this technology turns out to work particularly well.
True, Skynet may collect license plates and facial-recognition matches and camera feeds from all across the country; but videos must be watched, images cross-checked, and automated decisions second-guessed. In Wuhan, as Skynet’s output increased, the city was forced to hire over 2,700 employees to comb through the videos it produced; in Shenzhen, Skynet’s demands for 24-hour coverage required a district-level public security bureau to set up more than 170 new surveillance centers employing some 760 full-time police staff. The pitfalls of individual programs vary somewhat, but the broader story remains largely the same: again and again, Pei details the vast human infrastructure needed to support widespread technological surveillance at the scale of a country as large as China.
This ballooning of personnel to support technological surveillance programs mirrors the growth of a decades-old and millions-strong human informant network that has embedded itself deep into Chinese civil society. In The Sentinel State’s telling, many Chinese civic organizations are infiltrated to the hilt and heavily monitored by a panoply of state security forces.
In just a single city, informants successfully sabotage more than one hundred potential protests in a year; security agents attempt to ply a civil-society activist with gifts for her child; a lawyer in Beijing is tailed by uniformed police officers during politically sensitive anniversaries. Chinese academia, too, is rife with informants. Less elite schools, Pei observes, can sometimes get by with only a single tipster stationed in each department—not so at the prestigious Beijing Foreign Studies University, where the rate jumps to as high as one per classroom.
This resemblance between the ubiquity of the camera operator and the ubiquity of the informant demonstrates that the atmosphere of paranoia and restraint that sustains China’s dictatorship needs little in the way of advanced technology. “Simply knowing that secret police and their informants may be watching can lead one to curtail one’s anti-regime activities,” Pei laments. “Herein lies the true power of the Chinese surveillance state, [made] possible [by] millions of ordinary citizens watching and listening in the course of their daily lives.”
Only one part of this story typically gets told. In recent years, coverage of China has settled on describing a sort of mythical authoritarianism without authoritarians—the Chinese regime, we are told, endures and will endure because of unprecedented advances in technology that allow for never-before-seen forms of intrusion into citizens’ lives. Josh Chin and Liza Lin’s book Surveillance State alternates between descriptions of “digital utopia” and a “dystopian police state … armed with AI.” Elsewhere, Kai Strittmatter’s We Have Been Harmonized paints a picture of “digital totalitarianism,” according to the publisher’s description, where “facial and voice recognition, GPS tracking, supercomputer databases, and millions of high-resolution security cameras make it nearly impossible for a Chinese citizen to hide anything from authorities.”
It is against this discourse that Pei’s The Sentinel State positions itself. Pei’s argument is quite digestible: He contends that the endurance of autocracy in China is less a function of technology than it is of human labor in the form of informant networks, labor-intensive domestic spying, and heavy policing. Pei is not a denialist when it comes to Chinese authoritarianism, nor is he interested in downplaying the extent of China’s surveillance state. But he is keenly aware of the ways in which today’s Chinese autocracy looks more like the past than not, and he is wary of the extent to which this story of technological tyranny obscures important realities of the situation on the ground in China.
Though Pei does not make this case forcefully in his book—instead choosing to focus on the empirical details of China’s surveillance apparatus—The Sentinel State helps shine a light on how the idea of techno-authoritarianism serves as a kind of totalizing narrative: one which has pernicious effects on our ability to understand the political situation in China.
As stories about flashy new technologies eclipse more measured coverage, it becomes easy for foreign audiences, particularly those in America, to lose track of the actual harms inflicted by China’s surveillance state. Put more simply, this narrative of techno-authoritarianism takes a difficult-to-summarize situation filled with diverging interests and mundane realities and flattens it into the language of the American cultural imagination, inflected with lurid Cold War fantasy and as contemptuous of complexity as it can be.
Why has it taken so long for a book like The Sentinel State to emerge? If human surveillance explains the persistence of dictatorship in China, what explains the persistence of the techno-dystopian canard which keeps us from seeing this?
For starters, the myth of a techno-authoritarian China is deeply rooted in America’s cultural discourse. And this isn’t helped by the fact that the lines between futurist prediction and science fiction have become increasingly blurred.
In a country where informants permeate every layer of social interaction, the solution is not as simple as outsmarting an algorithm or evading capture on a video feed.
For example: Taiwanese computer scientist Kai-Fu Lee has sounded the alarm about an AI arms race between the United States and China in his 2018 book AI Superpowers. But he has also coauthored a 2021 English-language short story collection called AI 2041 with the science fiction author Chen Qiufan, another veteran of the Chinese tech industry. Like Chen’s 2013 novel Waste Tide—which features a high-tech Chinese system called Compound Eyes keeping watch over “every street, every corner, [and] every expression on every person”—AI 2041 gives us a flashy and entertaining view of a future digital dystopia.
Unsurprisingly, AI 2041 was a bestseller in the United States, with rave reviews in the Financial Times, the Wall Street Journal, the Washington Post, and the New York Times. This is not to fault Lee and Chen for writing a quite compelling work of literature. But it is worth noting that quasi-fantastic narratives of this type have found an eager audience in Americans, whose view of the reality of Chinese surveillance differs little from the fiction they read.
After all, “the world consumes much of its imagery of East Asia”—including China—“through pop cultural forms [in which] the region is largely depicted as a collection of high-tech, productive societies” with a “natural proclivity to master technologies of the West, from sedans to silicon chips,” according to historian Andrew Liu (leaning on the work of literary critic Colleen Lye). It is this imagined proclivity that “is precisely what makes Asia so dangerous” in the minds of Americans. And it is this idea—that China has taken the technologies of the West, and turned them authoritarian and malignant—that animates much of the narrative of digital dictatorship, which The Sentinel State works so hard to complicate.
This fascination with the role of technology in Chinese politics is continuous, in Liu’s view, with a broader cultural tendency in which “East Asia is either a hypermodern civilization that reflects the degradation of [the] US … or, conversely, [a] threat that will one day overtake Euro-America, subsuming it within itself.” Liu and other scholars dub this tendency “techno-Orientalism,” a trend that runs most obviously through the ubiquity of Asian megacorporations in genre-defining science fiction works like Blade Runner (1982) and Neuromancer (1984), and appears more subtly in The Matrix (1999), Ex Machina (2014), and other cultural products set in and around imaginary East Asias.
But techno-Orientalism is not just a problem of media representation, or even just of putative racism. Rather, it serves to obscure the true stakes of the debate over Chinese surveillance: subjecting the complexities of Chinese oppression to a sort of homogenizing American anxiety. In the techno-Orientalist frame, the problem is less about the danger the Chinese regime poses to its own people and more about the danger it poses to the West. For the techno-Orientalists, Liu observes, “decades of industrialization and accumulation in the Asia-Pacific region have brought to life [new] fantasies of an exceptionally dominant China [as] dystopian threat.”
This story also serves convenient political purposes. Putting technology at the center of the narrative about Chinese unfreedom allows American security hawks to frame China as a high-tech threat to the entire “free world.” And this, in turn, allows them to fold seemingly humanitarian concerns about surveillance into their broader project: China containment. The logic of the so-called US-China tech war shifts the conversation away from questions of human rights and toward questions of national-security policy, lumping the problem of the surveillance state in with debates about sanctions, semiconductor export controls, trade protectionism against Chinese electric cars, and the forced divestment of Chinese investors from apps like TikTok.
It doesn’t hurt, either, that this narrative places American commentators on the same footing as Chinese dissidents and activists, for whom the PRC’s surveillance state is a genuine threat. As China’s access to technology becomes the object of focus, it becomes easy to conflate the danger posed by specific technologies to victims of the Chinese regime with a vague notion of a tech-enabled global China threat, creating a narrative that both inflames nationalist paranoia in Americans and flatters the self-importance of America’s China hawks. China threatens me just as China threatens you, the story goes—so what’s good for me, an American defense analyst, must also be good for you, an exiled Chinese human-rights lawyer.
One of the most important observations in The Sentinel State comes near the middle of the book, when Pei remarks that “a society shot through with informants will be riven by distrust, eroding the foundations of democracy [and] social trust.” Pei is a political scientist and his book a recognizably social-scientific work of scholarship. Still, this observation about the long-term social effects of surveillance connects The Sentinel State to a long history of writing in the Chinese humanities, one which runs through the history of the People’s Republic, and perhaps even extends backward into imperial China.
Such a history can be seen in the Nobel Prize–nominated Chinese writer Can Xue, whose writing often returns to the subject of how Chinese society is structured by pervasive surveillance. Can Xue’s writing is deeply interested in how the basic form of Chinese social life begins to resemble the state’s apparatuses of supervision and control as trust erodes and suspicion proliferates. Can Xue’s characters live in a world of perpetual monitoring from which no one can escape, literary critic Jianguo Chen points out, but this monitoring comes not just from the “dictatorial party [but] the hateful masses as well.”
Take Can Xue’s novella Old Floating Cloud. Here, the line is blurred between state and social surveillance, between being watched by an informant and being watched by an inquisitive neighbor, or between watching someone and being watched. This blurring occurs as a character flits between observing and avoiding observation, gripped by a fear that his thoughts and feelings are as visible to his coworkers’ prying eyes as his neighbor’s actions are to his.
At home after work, he pretended to trim his beard. … With the mirror in his hand, he observed his neighbor’s movements behind him. He felt a little relaxed after assuring himself nothing suspicious was going on. Maybe his erratic heartbeat gave him away? … He thought everyone could hear the noise. That was why everyone at the office stared at him. … To prevent people from hearing his heartbeat, he sneaked into [the] office and looked out the window for hours [but] today, stretching his head out the window, he noticed heads at the other two windows [and] realized they were colleagues in the same room.
Conversely, in Can Xue’s Five Spice Street, we see echoes of The Sentinel State’s overwhelmed Skynet staff. Here, members of a neighborhood voluntarily, and with minimal coordination, create an atmosphere of “extensive surveillance,” in an incompetent attempt to catch what they believe to be sexual transgression. These neighbors thoroughly ruin the quality of life on their street. Yet, in the end, these self-appointed watchers cannot even agree on the appearance or age—let alone the behavior—of their target.
It is not by coincidence that one of the most incisive depictions of state and society in China comes from a literary figure. As noted by the critic and social theorist Nan Z. Da, close reading and other techniques of linguistic and literary analysis have long been key tools of Chinese surveillance. Such techniques date back to at least the Hundred Flowers Campaign, in which Mao invited intellectuals to candidly express their views of his regime before subjecting their writings to careful inspection for signs of punishable dissent. Chinese writers, Da observes, therefore have an understanding that surveillance can often take the form of “technical hairsplitting” or the “combing [of] text for minor, semantic transgressions.”
Coverage of Chinese civil society often applauds clever efforts to evade censorship using technological countermeasures, usually by way of coded phrases that censors cannot detect or encrypted messaging apps and video filters that allow Chinese netizens to avoid the attention of social media monitors. But if coded phrases and secure apps exist to evade digital censors, then an entire infrastructure of textual and cultural analysis—powered by humans and trained on a tradition of politically motivated close reading—exists as a sort of backstop for Chinese censorship. And, as Da notes, it has done so for at least decades.
In Hong Kong, for example, Chinese national-security authorities clamped down on protests since 2020. Consequently, cultural products—from novels to songs to children’s picture books—were scrutinized for signs of prodemocracy sympathies, in what the Hong Kong legislator Dennis Kwok decried as a “literary inquisition.”
This is all to say that—as both Chinese literature and The Sentinel State can tell us—there are specific harms perpetuated by China’s society of surveillance that cannot be seen clearly if we take something like “digital totalitarianism” as our lens of analysis. In a country where informants permeate every layer of social interaction, the solution is not as simple as outsmarting an algorithm or evading capture on a video feed. Rather, the threat of being watched by both state and stranger—of your writings and thoughts being turned over for dissent—shapes intellectual life in ways that technological surveillance by itself cannot.
Techno-authoritarianism may pit the dissident against the security camera. But an informant society pits the dissident against everyone, at least in their mind.
These are somewhat counterintuitive observations for many China watchers, but such insights about surveillance and social trust are by no means unique to Minxin Pei nor the writers and critics who can see them from the world of Chinese-language letters. They are simply reflections of the ugly realities of Chinese authoritarianism that a story of technological dystopia elides. Attempts to frame China’s surveillance state as a product of dastardly Oriental adaptation or a downstream effect of the wrong side winning the tech war obscure the true difficulty of the situation faced by Chinese citizens and human-rights advocates–—that the surveillance state cannot be sanctioned into nonexistence or tricked into irrelevance, because it lives in the heart of the country’s civil society and is inextricably tied to the form of its social relations.
The Sentinel State goes a long way to pointing us toward the above argument, because it recognizes that Chinese authoritarianism is fundamentally not a technical problem. And the book is useful because it not only reveals that the system does not rely on technology, but also that the harms it creates cannot be avoided by technical means. Pei’s ability to articulate this remains the strongest quality of his latest book: The Sentinel State returns us to the premise that politics is a human affair, and surveillance a human phenomenon.
Early in The Sentinel State’s description of the popular myth of Chinese techno-authoritarianism, Pei references George Orwell’s 1984, perhaps to lightly mock the most tired of all dystopian clichés. But whatever his intentions, Pei’s invocation of Orwell remains worth examining. Though 1984 gave us many of the images that still define tyranny for Western audiences—TVs watching their viewers chief among them—it ultimately isn’t these technological tricks that undo Orwell’s protagonist. Instead, it’s Winston’s misplaced faith in individuals who turn out to be part of a carefully oriented network of informants and secret police, the reveal of which undercuts his belief in the basic premises of social trust and destroys the last meaningful relationships in his life.
Even the most cited-to-death work of literary dystopia knows that technology can only do so much. At the end of the day, dictatorships are made up of people, their harms human-sized and their fortunes rising and falling with the currents of social life. 