The postponement of the stock market debut of AI company SenseTime was not newsworthy for Chinese state newspapers, such as the Global Times. That may be because of its connection to sensitive human rights issues: The US government had unequivocally linked the company to the total surveillance of the Uyghur people. Finn Mayer-Kuckuk analyzes the backstory of the entanglement of geopolitics, finance, and AI. The former flagship company has not only lost its glory, but also its innocence.
Beijing’s strict capital controls have driven casino operators in the Macau Special Economic Zone to be constantly on the lookout for new loopholes in recent years. Citizens of the People’s Republic can legally transfer only $50,000 outside the country per year. Macao’s casinos have thus lured wealthy people to their city with pleasure trips and paid them their winnings in US dollars or euros. The stakes had to be high enough – otherwise, access to VIP rooms was denied. With the arrest of gambling baron Alvin Chau, Beijing is sending a warning signal to the entire industry. Head of state Xi Jinping not only wants to bring the entertainment and technology industries back in line but also the financial and gambling industries, as our Beijing authors analyze.
Have a pleasant day!
SenseTime is one of the AI companies that already found a practical use for their technology. The company’s software provides facial recognition with unprecedented accuracy. But despite investor interest, the company has postponed a planned Hong Kong IPO for the time being. The reason is a decision by the US government. The US government has imposed new sanctions on SenseTime (China.Table reported), which also affect the company’s future financial position. The reason for the sanctions is the large-scale application of SenseTime products to monitor the Uyghur minority.
SenseTime is thus another renowned company that has become part of the political mix. However, there are differences to other companies affected by US sanctions. Telecom equipment maker Huawei, semiconductor manufacturer SMIC, or mobile phone operator China Mobile all point to having only superficial relations with China’s security authorities. Even though no one in China can escape the wishes of the state, the reasoning of these companies is solid. SenseTime, on the other hand, provides core components for the total digital surveillance state.
Last Friday was the fateful day for SenseTime’s stock market plans. On the one hand, December 10 is Human Rights Day. For US President Joe Biden, it also marked the end of a vaguely productive democracy summit to which China was expressly not invited. The US government then followed up with sanctions on the occasion of Human Rights Day and the Democracy Summit. SenseTime was a logical target because media reports had already created the necessary profile. Unlike previous US sanctions, this time it was explicitly in response to the oppression of the Uyghurs. Other sanctions were imposed on Myanmar and North Korea that day. The legal basis was the classification as part of the “military-industrial complex of the People’s Republic of China”.
Then on Thursday, rumors already surfaced that SenseTime could be blacklisted (China.Table reported). On Friday, SenseTime was supposed to announce details of the IPO. Instead, the management contacted the Hong Kong stock exchange and agreed to put the IPO plans on hold for the time being.
The nature of the US sanctions certainly affects the financial position of the company. Blacklisted companies are no longer obtain capital in the USA. On top of that, there is the resulting damage to the company’s image. On Saturday, SenseTime defended itself against the US Department of the Treasury’s reasoning for the sanctions. The accusations “are unfounded,” the company said. “We have complied with the applicable laws and regulations in relation to our business in all material respects in the jurisdictions where we conduct business.”
That is probably correct. In China, it is the one-party state that enforces a control regime based on its own laws. China could probably spin it the other way around and consider it as a criminal offense if it refused to participate in state operations. From an international perspective – and especially from the perspective of a value-driven foreign policy, which is once again being discussed these days – the role of the Shenzhen tech company should at least be viewed from different angles.
Throughout 2018 and 2019, it gradually became known how tight the surveillance of Xinjiang’s population had become. At the time, the New York Times published a series of influential articles detailing the use of modern technology in the region. According to the article, the software is able to distinguish between Uyghurs and Han Chinese based on their appearance. Police could also use this feature outside Xinjiang. For example, they allegedly also use software to extensively track the movements and actions of Uyghurs in coastal cities. According to reports, this software feature is referred to as “ethnicity recognition”.
This almost certainly made SenseTime of the companies that supplied the technical infrastructure for the establishment of a digital police state in Xinjiang. Along with other companies, SenseTime thus became synonymous with total surveillance in the eyes of the Western public. Then, in mid-October, the company announced it was seeking an initial public offering in Hong Kong. Immediately after the announcement, however, debates about its involvement in the construction of the surveillance state once again emerged (China.Table reported).
So SenseTime’s biggest advantage on the international stage is also its biggest problem: The Chinese state is the world’s largest possible customer for surveillance technology. Other countries also have competent providers of biometric recognition software, even if many of them are not quite playing in the same league as SenseTime. For example, the company Cognitec, whose facial recognition also works just as accurately, is based in Germany. It also cooperates with the German Federal Criminal Police Office (BKA). But the most notable project Cognitec has worked on was merely a test run for facial recognition in public spaces at Berlin’s Suedkreuz train station.
However, one camera placed above an escalator has already sparked a nationwide debate about the dangers of facial recognition. No contracts for large-scale personal recognition in public spaces are expected in Germany in the foreseeable future. An initiative to monitor 100 intersections by former Interior Minister Horst Seehofer failed miserably. Cognitec’s software therefore only sees internal use by the BKA for “photo research” in databases.
The Chinese providers such as SenseTime and Megvii, on the other hand, can further develop their software through a mass application. Government contracts bring the money in, which in turn flows into the products. The IPO in Hong Kong would have brought another windfall for SenseTime: The company planned to raise €680 million.
The initial interest of investors was not only connected to the Chinese state as the main buyer but also with its high level of technology. Like many AI providers, SenseTime uses neural networks for pattern recognition at the core of the system. This system uses a structure akin to a nervous system and learns key capabilities from samples. However, SenseTime also cleverly embeds pattern recognition within conventional software.
The software tracks people in the real world from camera to camera throughout the city. It also recognizes people not only by their faces but also by their gait, arm position, and other characteristics. To pick up these patterns, they only need to be captured once by a camera, together with the face. All this is what makes the software’s capabilities so valuable for total surveillance in Xinjiang, for example.
In 2018, SenseTime was already the highest valued AI startup in the world, with a valuation of $6 billion. By then, the company was already profitable. SenseTime is already working with hundreds of customers worldwide, including US chipmaker Qualcomm or Japanese carmaker Honda in the field of autonomous driving. Here, the system detects and evaluates what is happening in front of the car. Nevertheless, the company was already forced to downsize its IPO due to the human rights debate.
SenseTime was founded in 2014 by Tang Xiao’ou, a computer scientist from the Chinese University of Hong Kong. From the beginning, the focus was on developing their own algorithms. Many competitors use techniques that are freely available or simply the state of what is taught at universities. In the early years of the company’s history, SenseTime’s facial recognition, which worked particularly well, still caused enthusiasm. The start-up was seen as an example of how much more advanced China’s tech companies are.
The perception of SenseTime has already been permanently changed by the events of the past few years. The clear words of the US Treasury linking the company to human rights violations in Xinjiang now only add to the impression: Sense time has ultimately lost its innocence as a tech company. The company has lost the important segment of ethical or sustainable investors as a source of capital in its stock market plans. At the same time, the case of SenseTime shows that it becomes virtually impossible to please everyone by simply complying with regional laws in critical areas.
The COVID-19 pandemic has exposed how vulnerable global supply chains can be. The global supply chain disruptions caused by the pandemic have prompted policymakers around the world to urgently clarify how reliant their country is on imported goods and foreign suppliers. Some have already called for companies to reshore business activities back home to reduce the country’s (over-)dependence on foreign (rival) countries.
China also takes the (potential) threat to its economic self-sufficiency seriously. This is not only because China itself is directly suffering from the supply chain disruptions caused by the pandemic. As the world factory of almost all globally traded goods or/and their components, China is naturally interested in serving overseas markets without major obstacles. Since many firms in China still rely strongly on imported high-tech equipment and intermediates for their own production, China is also interested in ensuring a well-functioning supply of key components from overseas.
The indirect (potential) threat of the pandemic to the Chinese economy is relevant here as well. On the one hand, many, particularly Western, economies are strengthening their own industrial capabilities and capacities to ensure their technological sovereignty and economic self-sufficiency. This can further fuel the already existing economic and political tensions between China and Western economies such as the European Union (EU) and the United States (US) and further aggravate the difficulties Chinese firms face in sourcing key components from abroad. On the other hand, multinational enterprises (MNEs) affected by supply chain disruptions may relocate their business operations out of China. Since MNEs in China have been a key provider of financial resources, advanced knowledge, technologies and jobs, MNEs’ such relocation decisions could harm the Chinese economy.
These direct and indirect threats have alerted China to the increasing global competition for (financial) resources, for talents, for technology and for market as well as to the further increasing economic and political uncertainties Chinese companies are facing. Thus, in its recent 14th Five-Year Plan (2021-2025) for development, China emphasizes the critical importance to, firstly, promote its technological innovation capacity to ensure its self-reliance in science and technology and, secondly, promote its domestic economic circulation (as a core element of China’s dual circulation development strategy) to ensure economic self-sufficiency.
China plans to significantly increase its overall expenditures on research and development and to strengthen (Chinese) companies’ role in innovation. Both the transformation and the commercialization of new research results and technologies shall be scaled up. Technologically leading foreign companies and global talents shall be attracted to help close China’s gap to the technological frontier in critical technological fields.
China plans to explicitly strengthen its domestic economic circulation, i.e., boosting domestic demand and supply as well as ensuring demand, supply and upstream and downstream production stages along the supply chains to be coordinated more efficiently. China’s own industrial base shall be further strengthened and modernized. Its market size advantage shall be used to attract global resources for production and research. At the same time, the increased domestic supply of innovative and high-quality products shall help strengthen China’s role as a leading manufacturing and trading power.
In short, China goes for more domestic innovation, domestic production and domestic demand and it will be even more “selective” in promoting international trade and investments than in the past to help achieve its development goals. Chinese policies derived from the 14th Five-Year Plan are thus expected to strongly influence how future global supply chains will be reshaped. For example, MNEs with investments in or exports to China that may be considered by the Chinese government to hamper the Chinese firms’ development will face more difficulties in doing business as usual in China. And support for Chinese direct investments abroad can be geared to help enhance China’s own technological capability, thus helping upgrade Chinese firms’ position in the global supply chains in the long term.
China’s Five-Year Plan and its policies will therefore pose challenges not only for MNEs doing business in/with China but also for their suppliers and customers. This will further reinforce the need felt by policymakers from Western countries to in turn foster economic self-sufficiency and technological sovereignty of their own economies through decoupling.
At a time of considerable tension in the economic and political relations between China and Western countries, as well as increasing geopolitical conflicts and uncertainties, deglobalizing policy initiatives by China, the EU and the US may further weaken mutual trust, reduce willingness to cooperate, and ultimately lead to costly economic decoupling and less efficient re-nationalization of global supply chains.
To be sure, economic and political relations between China and the EU/US are complicated and there are economic and political conflicts of various kinds. Decoupling from each other is certainly not the best way to deal with these conflicts, however. Instead, to reap the (potential) benefits of globalization for global prosperity and sustainability, open bilateral and multilateral dialogues and cooperation are essential.
And for such dialogues, it is key not only to understand the interests of one’s own country, but also to learn more about the views of the other side. This is what the Global China Conversation # 5 “Reshaping Global Industrial Chains: Options for China” aims to contribute to.
Wan-Hsin LIU is a Senior Researcher in the Research Centers “International Trade and Investment” and “Innovation and International Competition” at the Kiel Institute for the World Economy. She is also Coordinator at the Kiel Centre for Globalization.
This contribution is prepared in the context of the event series “Global China Conversations” of the Kiel Institute for the World Economy (IfW). On Thursday, Qiyuan Xu, Deputy Director at the Institute for World Economics and Politics of the Chinese Academy of Social Sciences (CASS-IWEP), and Rolf J. Langhammer, Former Vice-President at the Kiel Institute for the World Economy, will discuss the topic: “Reshaping Global Industrial Chains: Options for China”. China.Table is media partner of this event series.
Artificial snow for the kicker: Preparations for the Winter Olympics have begun at the Shougang Big Air sports venue in Shijingshan. The big air freestyle skiing and snowboarding competitions of the 2022 Winter Games will take place there from February. The Big Air ski jump is located on the site of a former steel mill of Shougang Group.
The postponement of the stock market debut of AI company SenseTime was not newsworthy for Chinese state newspapers, such as the Global Times. That may be because of its connection to sensitive human rights issues: The US government had unequivocally linked the company to the total surveillance of the Uyghur people. Finn Mayer-Kuckuk analyzes the backstory of the entanglement of geopolitics, finance, and AI. The former flagship company has not only lost its glory, but also its innocence.
Beijing’s strict capital controls have driven casino operators in the Macau Special Economic Zone to be constantly on the lookout for new loopholes in recent years. Citizens of the People’s Republic can legally transfer only $50,000 outside the country per year. Macao’s casinos have thus lured wealthy people to their city with pleasure trips and paid them their winnings in US dollars or euros. The stakes had to be high enough – otherwise, access to VIP rooms was denied. With the arrest of gambling baron Alvin Chau, Beijing is sending a warning signal to the entire industry. Head of state Xi Jinping not only wants to bring the entertainment and technology industries back in line but also the financial and gambling industries, as our Beijing authors analyze.
Have a pleasant day!
SenseTime is one of the AI companies that already found a practical use for their technology. The company’s software provides facial recognition with unprecedented accuracy. But despite investor interest, the company has postponed a planned Hong Kong IPO for the time being. The reason is a decision by the US government. The US government has imposed new sanctions on SenseTime (China.Table reported), which also affect the company’s future financial position. The reason for the sanctions is the large-scale application of SenseTime products to monitor the Uyghur minority.
SenseTime is thus another renowned company that has become part of the political mix. However, there are differences to other companies affected by US sanctions. Telecom equipment maker Huawei, semiconductor manufacturer SMIC, or mobile phone operator China Mobile all point to having only superficial relations with China’s security authorities. Even though no one in China can escape the wishes of the state, the reasoning of these companies is solid. SenseTime, on the other hand, provides core components for the total digital surveillance state.
Last Friday was the fateful day for SenseTime’s stock market plans. On the one hand, December 10 is Human Rights Day. For US President Joe Biden, it also marked the end of a vaguely productive democracy summit to which China was expressly not invited. The US government then followed up with sanctions on the occasion of Human Rights Day and the Democracy Summit. SenseTime was a logical target because media reports had already created the necessary profile. Unlike previous US sanctions, this time it was explicitly in response to the oppression of the Uyghurs. Other sanctions were imposed on Myanmar and North Korea that day. The legal basis was the classification as part of the “military-industrial complex of the People’s Republic of China”.
Then on Thursday, rumors already surfaced that SenseTime could be blacklisted (China.Table reported). On Friday, SenseTime was supposed to announce details of the IPO. Instead, the management contacted the Hong Kong stock exchange and agreed to put the IPO plans on hold for the time being.
The nature of the US sanctions certainly affects the financial position of the company. Blacklisted companies are no longer obtain capital in the USA. On top of that, there is the resulting damage to the company’s image. On Saturday, SenseTime defended itself against the US Department of the Treasury’s reasoning for the sanctions. The accusations “are unfounded,” the company said. “We have complied with the applicable laws and regulations in relation to our business in all material respects in the jurisdictions where we conduct business.”
That is probably correct. In China, it is the one-party state that enforces a control regime based on its own laws. China could probably spin it the other way around and consider it as a criminal offense if it refused to participate in state operations. From an international perspective – and especially from the perspective of a value-driven foreign policy, which is once again being discussed these days – the role of the Shenzhen tech company should at least be viewed from different angles.
Throughout 2018 and 2019, it gradually became known how tight the surveillance of Xinjiang’s population had become. At the time, the New York Times published a series of influential articles detailing the use of modern technology in the region. According to the article, the software is able to distinguish between Uyghurs and Han Chinese based on their appearance. Police could also use this feature outside Xinjiang. For example, they allegedly also use software to extensively track the movements and actions of Uyghurs in coastal cities. According to reports, this software feature is referred to as “ethnicity recognition”.
This almost certainly made SenseTime of the companies that supplied the technical infrastructure for the establishment of a digital police state in Xinjiang. Along with other companies, SenseTime thus became synonymous with total surveillance in the eyes of the Western public. Then, in mid-October, the company announced it was seeking an initial public offering in Hong Kong. Immediately after the announcement, however, debates about its involvement in the construction of the surveillance state once again emerged (China.Table reported).
So SenseTime’s biggest advantage on the international stage is also its biggest problem: The Chinese state is the world’s largest possible customer for surveillance technology. Other countries also have competent providers of biometric recognition software, even if many of them are not quite playing in the same league as SenseTime. For example, the company Cognitec, whose facial recognition also works just as accurately, is based in Germany. It also cooperates with the German Federal Criminal Police Office (BKA). But the most notable project Cognitec has worked on was merely a test run for facial recognition in public spaces at Berlin’s Suedkreuz train station.
However, one camera placed above an escalator has already sparked a nationwide debate about the dangers of facial recognition. No contracts for large-scale personal recognition in public spaces are expected in Germany in the foreseeable future. An initiative to monitor 100 intersections by former Interior Minister Horst Seehofer failed miserably. Cognitec’s software therefore only sees internal use by the BKA for “photo research” in databases.
The Chinese providers such as SenseTime and Megvii, on the other hand, can further develop their software through a mass application. Government contracts bring the money in, which in turn flows into the products. The IPO in Hong Kong would have brought another windfall for SenseTime: The company planned to raise €680 million.
The initial interest of investors was not only connected to the Chinese state as the main buyer but also with its high level of technology. Like many AI providers, SenseTime uses neural networks for pattern recognition at the core of the system. This system uses a structure akin to a nervous system and learns key capabilities from samples. However, SenseTime also cleverly embeds pattern recognition within conventional software.
The software tracks people in the real world from camera to camera throughout the city. It also recognizes people not only by their faces but also by their gait, arm position, and other characteristics. To pick up these patterns, they only need to be captured once by a camera, together with the face. All this is what makes the software’s capabilities so valuable for total surveillance in Xinjiang, for example.
In 2018, SenseTime was already the highest valued AI startup in the world, with a valuation of $6 billion. By then, the company was already profitable. SenseTime is already working with hundreds of customers worldwide, including US chipmaker Qualcomm or Japanese carmaker Honda in the field of autonomous driving. Here, the system detects and evaluates what is happening in front of the car. Nevertheless, the company was already forced to downsize its IPO due to the human rights debate.
SenseTime was founded in 2014 by Tang Xiao’ou, a computer scientist from the Chinese University of Hong Kong. From the beginning, the focus was on developing their own algorithms. Many competitors use techniques that are freely available or simply the state of what is taught at universities. In the early years of the company’s history, SenseTime’s facial recognition, which worked particularly well, still caused enthusiasm. The start-up was seen as an example of how much more advanced China’s tech companies are.
The perception of SenseTime has already been permanently changed by the events of the past few years. The clear words of the US Treasury linking the company to human rights violations in Xinjiang now only add to the impression: Sense time has ultimately lost its innocence as a tech company. The company has lost the important segment of ethical or sustainable investors as a source of capital in its stock market plans. At the same time, the case of SenseTime shows that it becomes virtually impossible to please everyone by simply complying with regional laws in critical areas.
The COVID-19 pandemic has exposed how vulnerable global supply chains can be. The global supply chain disruptions caused by the pandemic have prompted policymakers around the world to urgently clarify how reliant their country is on imported goods and foreign suppliers. Some have already called for companies to reshore business activities back home to reduce the country’s (over-)dependence on foreign (rival) countries.
China also takes the (potential) threat to its economic self-sufficiency seriously. This is not only because China itself is directly suffering from the supply chain disruptions caused by the pandemic. As the world factory of almost all globally traded goods or/and their components, China is naturally interested in serving overseas markets without major obstacles. Since many firms in China still rely strongly on imported high-tech equipment and intermediates for their own production, China is also interested in ensuring a well-functioning supply of key components from overseas.
The indirect (potential) threat of the pandemic to the Chinese economy is relevant here as well. On the one hand, many, particularly Western, economies are strengthening their own industrial capabilities and capacities to ensure their technological sovereignty and economic self-sufficiency. This can further fuel the already existing economic and political tensions between China and Western economies such as the European Union (EU) and the United States (US) and further aggravate the difficulties Chinese firms face in sourcing key components from abroad. On the other hand, multinational enterprises (MNEs) affected by supply chain disruptions may relocate their business operations out of China. Since MNEs in China have been a key provider of financial resources, advanced knowledge, technologies and jobs, MNEs’ such relocation decisions could harm the Chinese economy.
These direct and indirect threats have alerted China to the increasing global competition for (financial) resources, for talents, for technology and for market as well as to the further increasing economic and political uncertainties Chinese companies are facing. Thus, in its recent 14th Five-Year Plan (2021-2025) for development, China emphasizes the critical importance to, firstly, promote its technological innovation capacity to ensure its self-reliance in science and technology and, secondly, promote its domestic economic circulation (as a core element of China’s dual circulation development strategy) to ensure economic self-sufficiency.
China plans to significantly increase its overall expenditures on research and development and to strengthen (Chinese) companies’ role in innovation. Both the transformation and the commercialization of new research results and technologies shall be scaled up. Technologically leading foreign companies and global talents shall be attracted to help close China’s gap to the technological frontier in critical technological fields.
China plans to explicitly strengthen its domestic economic circulation, i.e., boosting domestic demand and supply as well as ensuring demand, supply and upstream and downstream production stages along the supply chains to be coordinated more efficiently. China’s own industrial base shall be further strengthened and modernized. Its market size advantage shall be used to attract global resources for production and research. At the same time, the increased domestic supply of innovative and high-quality products shall help strengthen China’s role as a leading manufacturing and trading power.
In short, China goes for more domestic innovation, domestic production and domestic demand and it will be even more “selective” in promoting international trade and investments than in the past to help achieve its development goals. Chinese policies derived from the 14th Five-Year Plan are thus expected to strongly influence how future global supply chains will be reshaped. For example, MNEs with investments in or exports to China that may be considered by the Chinese government to hamper the Chinese firms’ development will face more difficulties in doing business as usual in China. And support for Chinese direct investments abroad can be geared to help enhance China’s own technological capability, thus helping upgrade Chinese firms’ position in the global supply chains in the long term.
China’s Five-Year Plan and its policies will therefore pose challenges not only for MNEs doing business in/with China but also for their suppliers and customers. This will further reinforce the need felt by policymakers from Western countries to in turn foster economic self-sufficiency and technological sovereignty of their own economies through decoupling.
At a time of considerable tension in the economic and political relations between China and Western countries, as well as increasing geopolitical conflicts and uncertainties, deglobalizing policy initiatives by China, the EU and the US may further weaken mutual trust, reduce willingness to cooperate, and ultimately lead to costly economic decoupling and less efficient re-nationalization of global supply chains.
To be sure, economic and political relations between China and the EU/US are complicated and there are economic and political conflicts of various kinds. Decoupling from each other is certainly not the best way to deal with these conflicts, however. Instead, to reap the (potential) benefits of globalization for global prosperity and sustainability, open bilateral and multilateral dialogues and cooperation are essential.
And for such dialogues, it is key not only to understand the interests of one’s own country, but also to learn more about the views of the other side. This is what the Global China Conversation # 5 “Reshaping Global Industrial Chains: Options for China” aims to contribute to.
Wan-Hsin LIU is a Senior Researcher in the Research Centers “International Trade and Investment” and “Innovation and International Competition” at the Kiel Institute for the World Economy. She is also Coordinator at the Kiel Centre for Globalization.
This contribution is prepared in the context of the event series “Global China Conversations” of the Kiel Institute for the World Economy (IfW). On Thursday, Qiyuan Xu, Deputy Director at the Institute for World Economics and Politics of the Chinese Academy of Social Sciences (CASS-IWEP), and Rolf J. Langhammer, Former Vice-President at the Kiel Institute for the World Economy, will discuss the topic: “Reshaping Global Industrial Chains: Options for China”. China.Table is media partner of this event series.
Artificial snow for the kicker: Preparations for the Winter Olympics have begun at the Shougang Big Air sports venue in Shijingshan. The big air freestyle skiing and snowboarding competitions of the 2022 Winter Games will take place there from February. The Big Air ski jump is located on the site of a former steel mill of Shougang Group.
