China’s Lead in Artificial Intelligence

China is leading the world in facial recognition technology, with companies such as SenseTime, iFlyTek, HikiVision and Dahua. Coined China’s “National AI Team” these companies are rolling out advanced features of artificial intelligence and facial recognition tools in the past decade. The move comes from Xi’s China’s move towards a “utopian” society, where safety is always guaranteed to citizens. As China aims to be the global leader in AI by 2030, it has already made ways in implementing its world-renowned technology in citizens’ daily life. Following its project to go “cashless” with the help of AliPay, the country has now started to use facial recognition for payment in selected spots.[1] People are now required to provide facial recognition data when buying mobile phones and registering new SIM cars, which is a policy that was implemented last year. This gives the government quite important personal information about the user connected to their facial ID. The same technology is being used in entry to residences, allowing for residents to opt out of using keys when entering their homes.

Social Score System

Xi’s China has vowed to use a social scoring system to track citizens’ behavior, and reward or punish them for their actions. Billboards on the streets will project the face and name of a person who is caught jaywalking in real time, which acts as a public shaming technique in the hopes of encouraging the person to “do better” in the future. This move towards “surveillance capitalism”, as coined by Harvard professor Shoshanna Zuboff, promises citizens that their safety is always ensured, and that actions that fall outside of the accepted Confucian social norms will not be permissible. The social score system that is now adopted deducts points from a person’s score upon detection of an undesired activity, such as spitting on the floor, being late on taxes or showing aggressive behavior. The digital presence and actions are tied to people’s ID, which can be investigated upon detection of their facial ID in a public space.

The objective was initially put to use with the “Sharp Eyes Project”, where a resident in a neighborhood was tasked with the role of managing the social scores of the citizens in the area. The assigned person would deduct and add points to the individual’s social score upon observation of a positive or a negative action, with the “aim of mobilizing […] the community”. This was the more “rudimental” application of the future of “Police Cloud”, as coined by the Washington Post.

The aim is to shape society into a servient one, where the societal system is applied and practised in a perfect manner. This falls in line with the country’s historical Confucian beliefs, where collectivism is valued over individualism. A low social score indicates that the person is not behaving “well”, which ends up in the government profiling the citizen as an “unsafe” one. This is supposed to be how criminals, or people with a tendency to act undesirably, are detected. For those who have watched it, the situation can be summed up by the world created in Black Mirror’s “Nosedive” episode.

Facial recognition technology’s leading firm SenseTime explained to VICE Media in a video interview that the technology is able to detect criminals in real time.[2] This helps safety by identifying people that may be deemed “unsafe” due to past activity in crowded places, after which the police are alerted and precautions are taken. This project is coined as the “Safe and Smart City” initiative. In a country of more than a billion people, this technology promises a lot for the peace of the country, as the chance of a criminal slipping through in a crowd of people is reduced to a slim one. Such advanced technology gives the government the tools to “finally achieve the level of control over people’s lives that it aspires to” have. As of 2020, 626 million cameras are equipped with the facial recognition technology, serving government’s aim is to “protect the legitimate rights and interest of citizens in cyberspace” by having the ability to track undesired activity that could harm societal peace by having access to all kinds of information and data that can enable them to act quickly and smartly.

However, all of this surveillance comes at a cost. This scenario somewhat resembles the dystopian world penned by George Orwell in 1984, where constant surveillance of the citizens is the main tool in “disciplining” society by an authoritarian government. The 24/7 CCTV presence in all corners of big cities denies citizens any privacy, as their actions are tracked and immediately linked to their social presence through facial recognition. The cost that is paid for the promise of ultimate safety is that no Chinese citizen can hide anything, and their data is collected at an alarming rate. The first lawsuit concerning this data privacy issue was brought up in November of 2018 by a professor at Zheijang Sci-Tech University, who claimed that this facial scan was taken without his consent at a zoo in Hangzhou.[3] Another point of concern was raised by parents at a university in Nanjing, where it was found that facial recognition technology to control students was installed without their consent.

While this is an important data privacy concern on its own, the real danger arises due to the categorisation that lies at the core of the social scoring system. The sociological effect of this system is that your social score determines your comfort in daily life. While those with higher social scores can enjoy benefits such as not paying deposits for homes and being treated with respect in public and private spaces, those with low social scores are prevented from buying plane or flight tickets, and are not allowed entry at specific locations. Though the identification of the people themselves is done through Artificial Intelligence, the categorization of the people is ultimately decided by humans. The three categories of “safe people”, “regular people”, and “unsafe people” may look easy to define at first, however the requirements and the checklists that lie beneath them take a dark turn when surveillance in the Uighur-majority Xinjiang region is investigated. The level of control that is exercised is open to use for any governmental policy, which is at the hands of quite loose regulation. This is why the consequences of a low score that await Chinese citizens of Han descent may not be promising, however the reality that awaits those in Xinjiang are far worse.

The Surveillance State of Xinjiang

Xinjiang, which is located in the northwest of China, is home to the Turkic Uighur minority of the country. An Islamic community, Uighurs are practicing Muslims who have been oppressed throughout Chinese history. The region was coined as “China’s Total Surveillance State” by the Wall Street Journal, due to the excessive use of AI and facial recognition to identify and oppress the minority public.

The repression of Uighurs dates back to the Mao era, where the Xinjiang era was taken over and into Chinese territory in 1949 with military force. The Uighur people’s Islamic presence was seen as a threat following Mao’s Religious Movement of 1958, where the communist government aimed to eradicate the practice of the religion. After that, religious education and practice was forcefully minimised to the point where the generations that followed the Movement were raised without awareness of their religion’s values and their ties to it. This was due to the immense control that was exercised over mosques and Islamic teachings in the religion. With the establishment and popularisation of the world wide web and the rise of globalisation, however, Uighurs were given a new perspective on their religion, and were given opportunities to explore and learn about Islam. With the rise of Islamophobia after 9/11 in the George W. Bush administration in the US, China, too, started to adopt an aggressive attitude towards their Muslim minority, seeing the region’s free practise of the religion as the “Talibanisation” of the Xinjiang. Xi’s reign in China saw a rise in Islamophobia, feeding the desire to further repress Xinjiang in aims to “integrate” the Uighur community into the Han Chinese style of living. With the ethnic riots that erupted in the region’s capital, Urumqi, in 2009, the government strengthened its efforts to change the region’s values and assigned Chen Quanguo to the region in 2016. Mr. Quanguo had previously served for the same purpose in Tibet, another region labelled as “rebellious” by the government. Quanguo adopted the policy of “transforming” Xinjiang, which meant that he aimed to turn them into “loyal China supporters.” The desire underlying this policy was to make Uighurs more subservient to the Chinese system. March 2017 saw the government passing a law banning and condemning “activities deemed manifestations of extremism”.

The activities that were deemed manifestations of extremism were never specified, which created the dangers that are posed by excessive use of facial recognition technology now. Facial recognition giant CloudWalk announced that they can recognise “sensitive groups of people” out of a crowd.[4] The categorization of people as “sensitive”, however, is completely dictated by the government, and is completely open to the dangers of racial profiling in a region as sensitive as Xinjiang. Yuan Yang of Financial Times states that “what counts as a terrorist in Xinjiang is ultimately decided by humans, not machines”.[5] This bears the question: What exactly does the government look for when they label people as “criminals” or “terrorists” depending on their low social score? Given the Islamophobic attitude of the Chinese government in the region, any show of religious activity by an Uighur citizen could put them at risk. As Xinjiang is surrounded by CCTV cameras equipped with FR technology, daily activities that might seem normal and acceptable outside of China could be used to deduct points from Uighurs’ social score, causing them to end up in a list of people deemed risky. As their daily lives are tracked at a scale unseen anywhere else in the world, the data that is collected from their phones, their digital activity as well as their live location gives the government the tools to oppress the people by labelling them as “suspicious”.

“The more powerful the tool, the greater the benefit or damage it can cause”[6]

The end result of this mass surveillance poses a threat to the well-being of the minorities. The internationally condemned “re-education camps”, which look to be more like modern day gulags, were created to “transform” such suspicious people. Though the government claims that such camps are only used to re-educate those with extremist views that could potentially turn into terrorist activity, the fact that the camps hold almost 10% of the Uighur population suggests something else. Currently, around 1 million are imprisoned to “make them forget about their identity”. There has been reports of torture, rape, forced consumption of alcohol and pork and other human rights violations in the camps, which has been made possible due to the constant tracking of the minority people. Minor things, such as reading the Qur’an, participating in religious teachings and sharing Islamic content on the Chinese communication app WeChat can cause you to end up in these camps. Victor Gevers, who uncovered an online database of personal information on the minority ethnicity, named the project as a “Muslim tracker” in his tweets. The fact that this is possible is very well known by the Xinjiang population, which has caused intense fear in the region. Recently leaked documents by the New York Times shows the government’s plans to extend restrictions in the region, and details how they aim to control the families of those whose relatives have been imprisoned, specifically students. Uighurs know that they are always being watched, and they are aware that FR technology makes it impossible for them to deny or hide their actions. The slightest mention of the government’s oppression can cause them trouble due to the constant surveillance, which is a useful tactic used by the authorities to silence the people. There are those that do end up being released from these camps, however their negative classification in society follows them everywhere. As all of their data has been tied to their facial ID, life is hard for them. A majority of public places require facial identification for entry, such as malls, mosques, restaurants and public bathrooms, and the system denies them entry due to their previous “suspicious” activity. In short, there is no escaping China’s demands of raising the Uighur identity.

A different way that the technology is used to “transform” the Uighurs is done through the technique of dilution of the minority population in the Xinjiang region. The region itself is resource-rich and also lies at a very strategic location. Xi’s China is known for its ambition to establish trade dominance not just in East Asia, but on a global scale. Alongside its heavy investments in South East Asia, Xi has also rolled out its “Belt and Road Initiative”, which is the re-establishment of the prestigious trade route called the Silk Road from a century ago. The Silk Road was known for spreading from Eastern Europe to China, and was deemed the main trade route for those looking to expand and practise their trade business. Xi’s “Belt and Road Initiative” aims to expand this historical route by establishing significant trade points along the way, which is predicted to extend all the way to Western Europe. This will be a critical move for China’s rise to global economic leadership, and the unrest in Xinjiang poses a great threat to it. The northwest location of the region is a primal point for the Belt and Road Initiative, as it provides access to Eurasia. The reason the government aims to “transform” the “rebellious” minority is to be able to use the region’s resources and location for its financial purposes. They have been doing this by diluting the percentage of the minority population in Xinjiang, as people of Han Chinese descent have been exported to the region as workers in the newly created businesses there. Bingtuan, named the Xinjiang Production and Construction Corp, is tasked with increasing the region’s population with China’s national ethnic majority.[7] This destroys the homogenous community that once existed in the region, posing a further threat to the peace of the people.

The FR database is fed with millions of data to achieve almost perfect ethnic identification and tracking, meaning the technology is capable of identifying Uighur people out of a crowd of those of Han Chinese descent. As Uighurs bear a more Central Asian look compared to their Han Chinese counterparts, picking them out from a crowd is quite easy with the help of the advanced technology.[8] This indicates that Uighurs can be identified and heavily tracked not just in Xinjiang, but in other parts of China as well. Claire Garvie summarises this by explaining that “if you make a technology that can classify people by an ethnicity, someone will use it to repress that ethnicity.” The data collected from all of the aforementioned devices is heavily used to track down Uighurs no matter where they are, and ensure that they cooperate with the government. The definition of “cooperation” is, again, completely up to the government itself.

Though this has been going on for quite some time, the issue started to gain international recognition, followed by condemnation in the last few years. This bears the question of what exactly China is doing wrong, how it is getting away with it and what should legally be done to eradicate the dangers that are posed to the ethnic minority.

[1]Feifei Liu, “Making Cutting-Edge Technology Approachable: A Case Study of Facial-Recognition Payment in China”, (Nielsen Norman Group, May 10, 2020) https://www.nngroup.com/articles/face-recognition-pay/

[2]Vice News,  https://www.youtube.com/watch?v=CLo3e1Pak-Y

[3]Celia Chen & Tracy Qu, “As facial recognition tech races ahead of regulation, Chinese residents grow nervous about data privacy”, (South China Morning Post, 27 November 2019),  https://www.scmp.com/tech/policy/article/3039383/facial-recognition-tech-races-ahead-regulation-chinese-residents-grow

[4]Isobel Cockerell, “Inside China’s Massive Surveillance Operation”, (Wired, 5 September 2019),  https://www.wired.com/story/inside-chinas-massive-surveillance-operation/

[5]Financial Times, “The role of AI in China’s crackdown on Uighurs”, https://www.ft.com/content/e47b33ce-1add-11ea-97df-cc63de1d73f4

[6]Brad Smith, “Facial recognition: It’s time for action”, (Microsoft on the Issues, 6 December 2019),  https://blogs.microsoft.com/on-the-issues/2018/12/06/facial-recognition-its-time-for-action/

[7]Jessica Batke, Mareike Ohlberg, “China’s Biosecurity State in Xinjiang Is Powered by Western Tech”, (Foreign Policy, 19 February 2020), https://foreignpolicy.com/2020/02/19/china-xinjiang-surveillance-biosecurity-state-dna-western-tech/

[8]Paul Mozur, “One Month, 500,000 Face Scans: How China Is Using A.I. to Profile a Minority”, (New York Times, 14 April 2019), https://www.nytimes.com/2019/04/14/technology/china-surveillance-artificial-intelligence-racial-profiling.html?campaignId=6WYWY