In Xinjiang’s Tech Incubators, Innovation Is Inseparable from Repression

In 2017, the Urumqi High-Tech Industrial Development Zone, in the capital of Xinjiang, published a catalogue of the “entrepreneurial and innovative” projects underway in its jurisdiction. The Xinjiang University Information Technology Innovation Park Co., Ltd. aimed to improve capacity for “grassroots makerspaces.” Another company was using blockchain “to establish an Asia-Europe big data trading platform.”

But the catalogue also heralded innovation of a very different kind. Also listed on the site was an “integrated grid management platform” that would “discover abnormalities, control the enemy’s social conditions, and maintain social stability.” Grid management, a state-imposed surveillance infrastructure that employs civilians to monitor their neighbors, exists across China. It divides cities and towns up into smaller units, with local neighborhoods serving as a mainstay of social control. In Xinjiang, however, where authorities have detained millions of Uyghur and other ethnic minority citizens, razed religious and cultural spaces, and otherwise criminalized indigenous identity, surveillance conducted through the grid management system has higher stakes. The Xinjiang government assumes the local populace contains “enemies” in need of extensive, and intrusive, monitoring. The “integrated grid management platform” featured in the catalogue would help grid management officers keep track of an intensive surveillance regimen of “visiting low-threat households once every two months, visiting basic households once a month, visiting key work households once every three days, and visiting key control households once a day” and “realizing precise positioning, precise personnel selection, and precise assignment of responsibilities.”

The High-Tech Zone’s grid management platform suggests the degree to which invasive surveillance and control systems have become embedded in Xinjiang’s efforts to promote technological innovation. Regional authorities praise projects sporting blockchain and O2O (online to offline) almost in the same breath as they commend technologies specifically designed to enhance control over ethnic minority populations.

Innovation and its benefits to society in Xinjiang have come to encompass both the use of big data to enhance cross-border trade and the use of big data to monitor people inside their own homes. Official documents promoting innovation in Xinjiang make no distinction between tools to help facilitate ethnic and religious repression and those designed to advance good governance. The broad use of the term “innovation” to embrace such seemingly incompatible intentions also reveals why they are not incompatible at all: To the Party-state, generating efficient, high-tech ways to control the behaviors of religious and ethnic minorities is innovative, it is modern and progressive, it is a moral good. Both domestic and foreign companies choosing to engage with the Xinjiang authorities on “innovative” projects and technologies should understand this mindset, even if the collaborative projects themselves have no obvious connection to repression in the region.

Two resources offer insight into Xinjiang officials’ definition of “innovation.” The first is China’s national “entrepreneurship and innovation” website, the online hub for the country’s larger effort to modernize its economic system and boost job growth. The site contains information on 120 “demonstration bases” around the country, including the High-Tech Zone in Urumqi. The second resource, the website for the Xinjiang Science and Technology Achievement Management Network, lists projects and organizations to which the government has awarded science and technology prizes.

The 2017 Urumqi High-Tech Industrial Development Zone work plan listed projects included in its “entrepreneurship and innovation support system.” The first six projects on a list of 60 included a “local color” town, described by Xinjiang’s official news website as a “cultural tourism” site designed to highlight local “folk customs”; a platform to help organize information gleaned from reports by government officials assigned to live with and monitor ethnic minority families; a “one-stop” government procurement platform; a “low-altitude drone industry service system” integrating research, training, maintenance services, and big data; a “Central and Western Asia gene-detection technology application demonstration base”; and a “Geekspace” providing “a new type of low-cost, convenient, total-factor makerspace for geeks.”

The first six projects listed in the 2017 Urumqi High-Tech Industrial Development Zone work plan’s “entrepreneurship and innovation support system”:

While some of these projects, like the “Geekspace,” seem of a piece with tech incubators worldwide, others facilitate surveillance of local ethnic minority populations. The “Visit the People, Benefit the People, and Bring Together the Hearts of the People” platform, designed to help cadres manage information gleaned from their long-term stays in predominantly ethnic minority areas, stands out as a technology that enables exceptionally intrusive government surveillance. Other efforts, like the low-altitude drone system or the project to sequence genes specific to Uyghurs and other central Asian populations, also have potentially repressive applications, given the regional use of both technologies to surveil or identify individuals targeted by authorities. Even the “local color” town likely entails forms of coercion. While the Chinese government has ramped up its efforts to stamp out genuine expressions of Uyghur culture in recent years, it has also boosted propaganda showing, as The New York Times described it, “Uyghurs and other minorities . . . singing and dancing happily in colorful dress” as a means to hide the grim effects of its policies from both the outside world and from Han Chinese citizens. “Anningqu Local Color Town” likely serves yet another effort to define, standardize, control, and commodify indigenous behaviors—the sort of project one expert has called the “museumification” of Uyghur culture.

Likewise, on the Xinjiang Science and Technology Achievement website, the government lauds “innovative” projects that treat coal and chemical industry wastewater alongside those that assist cadres surveilling people in their own homes, such as the “Ethnic Unity One Family” campaign, which mandates that government employees stay in the homes of assigned Uyghur charges. Unsurprisingly, given the extent of state monitoring of Uyghur and other Turkic populations, the site also contains quite a few projects focused on video surveillance analysis software.

The Science and Technology Achievement website also celebrates a number of language-focused projects. In 2018, the site commended “a multilingual, real-time voice translation device” to assist Mandarin-speaking cadres stationed in Xinjiang. Though the cadre homestay programs had successfully “strengthened ethnic exchange and fusion,” in southern Xinjiang, “an area inhabited by ethnic minorities,” language barriers made it “difficult to interact with the ethnic minority masses or accurately communicate the Party’s policies.” Essentially, the project involved developing a gadget that allowed government minders living in residents’ homes to more effectively tell them how to behave.

In 2016, an article on the website extolled a suite of “key technologies for content monitoring on targeted Uyghur-language websites” that would help “strengthen counter-terrorism and stability maintenance, and battle against the forces of domestic and international religious extremism, the forces of ethnic separatism, and the forces of international terrorism” (what the government calls the “three forces”). Advanced technologies, both foreign and local, would be needed to provide a “scientific safeguard” against terrorist and extremist threats. Specifically, this meant developing software that could analyze written Uyghur text, including name and emotion recognition capabilities. Similarly, another 2018 project aimed to develop a “multilingual public opinion information monitoring system” specifically because the “three forces” had “ideologically infiltrated our region” through the use of “minority languages.” Both of these projects took as their premise that Uyghur and other minority languages are a key threat vector.

In some cases, it can be impossible to disentangle the different strands of tech-enabled social control woven into a single project. In 2020, the Xinjiang Academy of Agricultural Sciences, along with other research institutes, developed a technique to prevent certain types of disease among the region’s cotton plants. To more widely disseminate this finding, “456 trainings were given to a total of 94,000 people (including through the ‘Visit the People, Benefit the People, and Bring Together the Hearts of the People’ program, villager night schools, and ‘relatives’ embedded in locals’ homes).” That is, authorities utilized existing coercive mechanisms—some of which were themselves abetted by “innovative” technology—to teach Uyghurs and members of other ethnic minority groups how to improve cotton yields. That this particular project revolved around cotton makes it all the more egregious; the region’s local authorities have a long history that continues today of forcing Uyghurs to pick cotton against their will.

To a certain extent, the Chinese government’s lack of differentiation between seemingly neutral and obviously invasive technology accords with how governments the world over think about and interact with technological innovation. Cities throughout the United States have snapped up technology purporting to help stop or prevent crime, without fully appreciating (or even much caring) about the very real privacy and human rights implications of their purchases. Individuals, too, engage with new technologies without much thought about how those same technologies might be used in nefarious ways. An Apple AirTag may keep a person from losing their car keys, but it could also allow someone else to track their movements at a disturbingly granular level. As Darren Byler, who studies technology and politics, puts it, “information infrastructures make the intimacy of social behaviour and daily movement available to the gaze of the state and technology companies.”

In the context of Xinjiang, however, such “innovation” supports a much more intentionally crafted ecosystem of repressive policies, infrastructure, and technology. The Party-state has deployed cutting-edge technologies from the research and commercial sectors to fashion an unrivaled system of population monitoring, assessment, and control. It has rewarded companies and universities that create products closely aligned with its goals, welding together more benign notions of “progress” and “innovation” with the goal of re-making the population of Xinjiang in its desired image.