Since Xi Jinping took power in 2012, China has become a less open country. Stricter rules and regulations adopted during his presidency, coupled with a more centralized system established after the March 2018 constitutional reform, changed the social and political context, allowing the government to suppress social discontent and increase government control over the population. A control that is exerted more and more through technology, pushing to the extreme what has been labeled as the age of “surveillance capitalism”, that is, the use of artificial intelligence (AI) software and machine learning tools that study consumers’ behavior in order to predict their likes and preferences.
However, the new measures the Chinese government is adopting should not come as a surprise, since they have constantly characterized Chinese history in terms of government control over society. For example, during the Warring States era (481-221 B.C.), Chinese states adopted a specific household system, which facilitated spying activities from the central government as well as among the population. During the Song dynasty (960-1279 A.D.), a similar, more sophisticated system was put in place, the so-called baojia system (保甲制 baojiazhi): a community-household system through which the central government could pervasively control the local population even through programs of collectivization. The baojia system survived, with several modifications, until today with the establishment of the social credit system, a government program, which measures citizens’ behavior and distributes scores and other welfare-like benefits according to their social performance.
Surveillance measures that have become even more pervasive, especially in relation to a series of anniversaries of past historical events that shaped contemporary China. Beijing celebrated at the end of April 2019 the one-hundred-year anniversary of the May Fourth Movement; an intellectual, anti-imperialist movement, which in 1919 protested around the streets of Beijing against the conditions of the Treaty of Versailles – signed at the end of the First World War – and against the weak, newly established Chinese republican institutions, accused of also being partly responsible for that detrimental war outcome, which further weakened China at the world stage. At the same time, 2019 represents the anniversary of two other more dramatic events that critically damaged the image and the domestic stability of China: the 1989 Tiananmen massacre – conducted by the Chinese government against students gathered in Tiananmen square asking for more democracy – and the suppression of the 2009 Xinjiang riots, conducted against the Uighurs – the ethnic Muslim minority inhabiting Xinjiang, a crucial Chinese province located in the northwestern part of the country.
Since 2009, the Chinese government has developed new technological tools to strictly control the people of this province. As a response, the East Turkestan Islamic Movement (ETIM) conducted two terrorist attacks outside Xinjiang: the 2013 Tiananmen and the 2014 Kunming railway station terrorist attacks.
After Xi Jinping’s 2014 new laws on anti-terrorism and the promotion of the so-called Strike Hard Campaign against violent activities and terrorism – 严厉打击暴力恐怖活动专项行动 yanli daji baoli kongbu houdong zhuanxiang xingdong – symbolized by the slogan “people’s war on terror,” China has heavily relied on technology and machine learning algorithms. The objective is to make sure that security is controlled, monitored and analyzed less by manpower than technological means. However, people – in this case police officials – are still quite relevant for data collection, as highlighted by the use of fanghuiju teams (访惠聚). Their name refers to the acronym: Visit the People访民情, Benefit the People惠民生, and Get Together the Hearts of the People 聚民心.
Since 2014, one of the first and clearest manifestation of the Chinese control over the Xinjiang’s population has regarded the creation of real concentration camps – defined by the Chinese authorities as “reeducation and training” camps – where many Uighurs are detained. In those confinement camps, adult males are forced to listen to hours of Communist Party propaganda, renounce the reading of the Koran, eating pork, and drink alcohol. Through these social engineering methods, the Chinese government objective is to modify the local population culture and religious practices. Sending thousands of Chinese people – belonging to the Han ethnic group – to Xinjiang is also part of this overall project, since this would also facilitate the ethnic transformation of the region.
A turning point on this relocation policy happened in 2017, after the appointment of Chen Quanguo as the new governor of Xinjiang. Having been acknowledged for his successful measures in repressing protests in Tibet, he applied in Xinjiang the same methods adopted before by developing the so-called grid management system (网络化治理体系 wangluohua zhili tixi), which relies on a capillary system of territorial surveillance.
The application of these physical installations, coupled with the development of AI and machine learning, has allowed China to conduct racial profiling, hence becoming the first country to implement what has been defined as an automated racism.
Specifically, the Chinese government has created the so-called Integrated Joint Operations Platform (IJOP 一体化联合作战平台 yi tihua lianhe zuozhan pingtai), a regional data system that uses AI to monitor the countless checkpoints in and around Xinjiang’s cities. Surveillance cameras around the country have been set up in a way to immediately identify Uighurs – their social profile and their facial attributes – around the streets. This technological software is easy to implement thanks to Uighurs’ clearly different facial traits with respect to the Han ethnicity, since most of the Uighurs look more like Central Asian people. This IJOP could immediately warn the police if any suspected Uighur enters any public institutions, such as banks, hospitals, shopping malls, and parks. At the same time, the AI programs can also target sensitive groups of people, by warning the Chinese government if an increasing number of Uighurs tend to gather in a specific area in a short period of time, therefore increasing the suspects of a possible terrorist planning.
Another technological measure aimed at tracking Uighurs movements around the region and the main cities, also conceived as a means to counter possible terrorist attacks, regards the installation, since February 2017, of the Beidou satellite navigation system in all the vehicles. At the same time, the central government has also targeted butchers and restaurants. Since most of the recent terrorist attacks, which occurred in Xinjiang, had been carried out by using knives and blades, butchers and restaurants had been forced to engrave on every blade the QR code containing the owner’s identity card information. In addition to that, in the last couples of years the government has also forced local population to download two apps into their phones – Jingwang Weishi and Baixing Anquan, meaning respectively Web Cleansing and Citizen Security – as an additional means to control and collect data among the population.
Finally, other relevant AI programs that the Chinese government is putting in place regards the so-called “health check,” which is applied to all adults of the region. It involves the collection of biometric data, such as DNA, blood type, fingerprints, voice recordings and face scans.
It is relevant to point out that all these technological measures respond to the national strategic program of the civil-military fusion – 军民融合junmin ronghe. Integrated joint operations represent the new People’s Liberation Army doctrine. It relies on a hi-tech C4ISR (command, control, communications, computers, intelligence, surveillance, and reconnaissance) “system of systems.” Consequently, such a technological approach facilitates the application of military models for police surveillance. China’s Ministry of Public Security has financed for billions of dollars two government plans – Sharp Eyes and Skynet – to computerize surveillance and facilitating intelligence collection for both police and military use. Through those two programs the Chinese government has also encouraged and financed start-ups to develop sophisticated systems for facial recognition and surveillance. The major Chinese AI companies involved in this project concern SenseTime, Yitu, CloudWalk, Megvii, and to a lesser degree Hikvision, whose core business is to sell cameras instead of creating facial recognition software. Most of the products these companies sell are Dragonfly Eye, Fire Eye, and Sky Eye, whose major function is to rely on AI to analyze footage from China’s surveillance cameras.
For what concerns IJOP, Xinjiang Lianhai Cangzhi Company (新疆联海创智公司) is the firm supporting the system. This company is a subsidiary of China Electronics Technology Group Corporation (CETC 中国电子科技集团公司), a major state-owned military contractor in China.
Hence, Chinese technological development and its application within its own borders is becoming one of the most debated topics of the current international affairs, since it raises many questions not only on the institutional future of the next likely superpower, but also on the nature of its technological devices and how pervasive they can be at the international level. China’s desire to control its territory by relying on technological tools also responds to its quest for international prestige. It should be kept in mind, in fact, that Beijing is exploiting the Xinjiang surveillance project in order to sell its technology to African countries, Zimbabwe being a case in point, where the central government is adopting Chinese technology for domestic surveillance. Another relevant example regards the highly influential annual China-Eurasia Security Expo in Urumqi, in Xinjiang through which Beijing sells military and technological equipment to many different countries. Another attempt at portraying itself as the champion of technological development, notwithstanding the current trade war with the US.
The views, opinions, and thoughts expressed in the text belong solely to the author, and not necessarily to the author’s organization.