By Tsunghan Wu

According to the Financial Times in December 2019, a leaked source reported that several Chinese telecommunication companies including ZTE, Dahua Technology, and China Telcom are proposing a draft for a new international standard at the International Telecommunication Union (ITU) which, if passed, would apply to facial recognition system, camera equipment, and resemblance surveillance tools in the future. [1] While discussion over the proposal is still ongoing as of the time of writing (May 2020), the impact that the new potential standard may cause, has drawn concerns from activists and governments in the West and like- minded democratic countries.

Two implications are considered here. For one thing, China has been explicit in its ambition and confidence with its growing capability to be a rule-maker on the global stage. The releases of Made in China 2025 and China Standards 2035 disclosed these blueprints and acts.[2] For another, the event itself reflects that the UN and its associated organizations have been targeted by the Chinese authority as effective platforms to promote and shape its image and power, which furthermore arguably may transform the current global order. This essay argues that the Chinese submission of the aforementioned proposal presents a vivid case of such, against a context of Sino-American strategic competition in recent years.

From below, this essay offers a brief review on facial recognition technology and analyzes China’s dynamic and salient achievements in this area. It also analyzes China’s recent commitment at the ITU.

Facial recognition technology can be used to verify or identify a person through a pre-established database. It is a sub-sphere of biometrics and does not refer to a specific method. Broadly, multiple approaches include geometric shape-based, local face-based, eigenface-based, neural networks-based, etc., can reach to the same goal while adopting different algorithms. This technology is strongly related to areas of telecommunication, artificial intelligence (AI), machine learning, and deep learning. It has been pushed forward by a set of scientific experiments and an interdisciplinary field of research where computer scientists, social scientists and governmental agencies have engaged with time and space.[3]

Facial recognition technology was pioneered by Woodrow Wilson Bledsoe, Helen Chan Wolf, and Charles Bisson in the US in the 1960s, then with advances in the following decades by other groups of scientists, including Larry Sirovich, Michael Kirby, AlexPentland, and Matthew Turk. During the time, the US Defense Advanced Research Projects Agency (DARPA) under the Department of Defense and National Institute of Standards and Technology (NIST) became increasingly involved in the development and commercialization of the technology.[4] Given this trend, facial recognition soon appeared in other regions.

Since the 2000s, development of facial recognition technology has been further aided by Big Data by not only many national governments around the world, but also enterprises and social media companies. These institutions deployed relevant systems on various occasions on various devices for various purposes. Governments including the US, some member states in the EU, the UK have usually adopted this technology to tackle criminals and terrorism, in particular in the post-911 context.[5] In 2011, the US Army used it to confirm the death of Osama Bin Laden.[6] On the other side, governments such as China also began adopting this technology under the same banner.

Facial recognition technology comes with many benefits indeed. Combined with other devices, it can be widely employed for users to access to their mobile phone or working/residential environments etc. Law enforcement agencies also use the technology to prevent crime and violence by deploying it on public occasions.

Taking a most recent example, it has been reported that facial recognition has been helping to curb the COVID-19 epidemic in China amid the worldwide outbreak. The government has deployed the system combined with other sensors to detect people with abnormal body temperature. With this system integrated in the nationwide CCTV network, the government can trace travel history of individuals.[7]

But where there are advantages to facial recognition technology, its shortcomings have similarly at a large degree drawn attention. Activists suggest that the technology, if abused, will pose challenges and even threats to personal privacy and human rights, causing ethical concerns. The line between public and private life is then blurred. Furthermore, authoritarian and totalitarian regimes may crack down on opponents by utilizing the technology to strengthen civil and social surveillance. It goes without saying that there is an issue about data protection and security. As a result, how to balance the technology’s brought benefits and negative aspects, meaning the convenience and privacy breach deserves exploration.[8]

To discuss the technological development of facial recognition, as well as the broader fields of telecommunications and AI, China is a subject that is unavoidable, as the country has become a driver in the relevant areas. Chronologically speaking, it is a latecomer, but in terms of business scale and innovation capability at this time, it is definitely a great power. In general, China built its facial recognition technology under its national development programs in telecommunications and AI, starting in the 1990s. With the rapid rise of overall state socio-economic capacity, China reached leapfrog progress in the 2000s. This section presents relevant policy programs and white papers guiding Chinese evolvement, and it focuses on the era of Xi Jinping.

Chinese development usually maintains a state-led character, and this case is no exception. Having understood the significance and power of telecommunications and AI, China believes that developing such technologies is necessary in order to prevail in the global power competition for the future. Although the leadership often sets frameworks dictating the industrial direction, it has always granted some flexibility for the business sector. This ping-pong interaction between the state and society has resulted in mutual reinforcement over telecommunication and AI technology; since President Xi Jinping took office, China’s aggressive national programs have brought about great progress within a short time.[9]

On May 8, 2015, the State Council of China launched the Made in China 2025 program, serving as a guideline to lead China towards becoming a global leader in innovation in manufacturing within the next thirty years, with the transformation of the current industry into an AI-based industry as its core strategy. [10] Two months later, another order on internet development by the State Council was published, vowing to cast a greater investment in and strengthening of digitalization in industry, economy, and social life. In particular, AI has been highlighted and the order sought to make AI technology and Internet development mutual reinforcement. [11] Following this, National Development and Reform Commission under the State Council claimed in May 2016 that it will accomplish the creation of large public databases of video, audio, image, map, etc., for AI Deep Learning by 2018. National standardization of the AI industry and the fostering leading-edge AI enterprises at global level were set as major goals.[12]

In 2015, the Chinese government again emphasized the strategic significance of the AI industry in its 13th Five-Year Plan (2016-2020), one of the most important national programs every five years.[13]

In July 2017, the Chinese government published a white paper targeting AI development. The importance of this white paper was due not only to the fact that it narrated a strategic insight as to why China must become involved in the global AI competition, but also the inclusion of a provision with a comprehensive proposal for how the government and the private sector can work together. The specific idea of a civil-military fusion stood out in this context. In addition, the white paper also set a goal for China to build an independent supply chain and become an international standard-maker by 2030.[14]

In 2018, the government announced 56 state-sponsored projects for the internet, AI innovation, and the digital economy, in which facial recognition technology was specifically upheld as a main area.[15]

In the same year, the Standardization Administration of China under the State Council published a white paper on the AI industry, which encouraged more investment and study in the field, as well as cooperation among enterprises. The white paper also advocated for accelerating standardization in China, paving a way towards entering the international market.[16]

To give a concise picture, the Chinese telecommunication technology has sharply upgraded since the Made in China 2025 program initiated. Then, having seen a recent release of China Standards 2035, Beijing showed a more active gesture in involving in international standards-making process. Massive funding has spurred business and research, which in return has strengthened overall Chinese performance in international competitions. The Carnegie Endowment for International Peace published the AI Global Surveillance (AIGS) Index in 2019, by which it notes that 52 out of 75 states with AI surveillance tools, or about 70%, have deployed Chinese facial recognition systems. [17] In total, 63 countries are using Chinese technology-based AI monitoring systems. Recently, the Nikkei Asian Review has also noted that China surpassed the U.S. in the number of international patent applications in 2019, emphasizing the importance of telecommunications innovation to Chinese decision-makers. [18] The top 50 companies working on facial recognition technology in China are listed in the end of this paper (See Table 1).

This section has addressed China’s accumulated abilities. In next section, we turn to the analysis of the Chinese proposal for facial recognition at the ITU in December 2019.

The ITU is an organization under the UN responsible for services related to the internet, telephone, radio, satellite, etc. Membership includes officials of UN member states, academic institutions, enterprises, and various public and private “professional units”, specializing in traffic, finance, healthcare, etc. As an international organization, however, the ITU has no enforcement powers, and can only request its members follow or adopt “international” standards. In fact, the standards are officially referred to as “recommendations” that they merely function as reference for any individual units establishing standards internally. Even so, given the ITU’s widespread presence, it must be a field where Beijing aims to expand its influence.[19]

Prior to the Chinese submission of its proposal to the ITU, the second meeting of the Biometric Characteristic Technology Commission under the China National Information Technology Standardization Commission was held in Beijing on November 20, 2019. During the meeting, 27 enterprises, including Tencent, Sense Time, Ping An, Ant Financial, Dahua, iFLYTEK etc. altogether announced the establishment of the National Standardization Working Group for Facial Recognition. According to Xinhua, this Working Group will not only formulate national standards in China, but also promote such Chinese standards onto the international arena. Chronologically speaking, the Chinese proposal at the ITU may be a follow-up commitment.

Promoting the Chinese standard up to an international one will also greatly benefit Chinese enterprises dominating the global market. If this comes true, for one thing, Chinese companies will be able to access to the market more easily compared to having to adopt other standards. While traditional economics theories claim that monopolies often lead to inefficiency, the success of digital giants such as Google and Amazon in this era implies that the wider and larger collection of data, the more powerful and efficient their algorithms and services. Eventually, they seem to enjoy incomparable advantages.

For another, those adopting different standards would face great challenges. Firstly, they might lose customers. In addition, if they are going to adapt into the Chinese standard, they will have to set up new production lines and even pay patent fees to Chinese companies. Costs of the adapted companies under this situation must increase. In either case, it is quite possible that enterprises not adopting the Chinese standard would fallen.

In June 2019, a standard on intelligent lamps was passed at the ITU identical with the standards of ZTE’s products. As a result, ZTE’s products should enjoy more favorable conditions compared to its competitors.

Importantly, China may achieve a breakthrough amid its contestation and confrontation with the U.S. in their trade and technology war since 2018, driving new progress. From this perspective, the Chinese submission of its proposal to the ITU at this time seems to be an attempt to take a stand over issues such as Huawei 5G technology. As the US-led group of countries regard Chinese products with suspicion over data security or privacy violations, setting the international standard will help justify the Chinese standpoint.

Chinese facial recognition technology has been known for its advanced quality along with its competitive cost. However, the Chinese product is often criticized in terms of its wide application to its surveillance tools, which have not only widely been used in the minority regions of China, but also the whole of Chinese society. Since the revelation of “reeducation camps” in Xinjiang, human rights campaigners have paid great attention to the abuse of surveillance.[20] The website Compritech published two reports in August and December 2019 respectively, highlighting that China reportedly performs poorly with the protection of biometric information from the city- to the national-level.[21] Since December 1, 2019, Chinese authorities have implemented a new policy of real name registration, demanding all new telephone users to complete facial registration and submit personal information.[22]

In past cases, ITU recommendations were listed in a form of guidelines without much in the way of details. However, the Chinese proposal presents details where possible, especially with regards to which biometric features should be permissible to collect and how this technology can be applied. Such features are therefore linked to Chinese characteristics.

To deal with the international concerns and criticism, Beijing has raised its own narrative, insisting on the usefulness and necessity of using facial recognition systems. It is true that not only China, but also many Western countries have deployed the technology nationwide for various purposes, and such has then led China to stand firm in justifying its massive usage of the technology. In addition, with its great success with Chinese products entering international markets over the past few years, Beijing has become more confident and active in engaging in international standard-making, which is also aligned with the Chinese national programs Made in China 2025 and China Standards 2035. Moreover, Beijing has attributed its alleged control over the COVID-19 epidemic to its policies of containment, within which the Chinese surveillance tools have played a visible role in a time when most countries are still suffering from the epidemic. To some extent, Beijing aims to shape a public opinion about the superiority of the “Chinese model,” be it its institutions, value, or products.[23]

However, there is no thought more chilling than “Big Brother is watching you,” a concept proposed by English writer George Orwell in his classic novel 1984, depicting a surveillance society.[24] The Chinese argument, in fact, ignores the differences in the scale and scope of data collection and applications between China and the West. By doing this, it seeks to dilute ethical and privacy concerns. Also, the success of Chinese products should be attributed to little more than its competitive pricing. And speaking of its control over COVID-19, which has been one of most serious global disasters since World War II, Chinese containment policies have been successful and inspiring, but the controversy over the violation of privacy and human rights is still concerning. In addition, there are many outstanding development models in the rest of world, with the Taiwanese model being one of them.

For reasons unknown, despite the ITU planned to complete this process by December 2019, the Chinese proposal is yet still under discussion. The Chinese attempt has been revealed, and those who are concerned about the issue are expected to keep their eyes open and possibly take actions.

Tsunghan Wu is a postdoctoral research fellow at Institute for National Defense and Security Research (INDSR), Taiwan. He received his PhD from King’s College London, and M.A. from National Taiwan University. His research interests include politics of telecommunication technology, nationalism, and ethnic conflict.

Table 1: Top 50 Companies on Facial Recognition Technology in China