Emerging technologies will play a major role in great power competitions in the 21st century. As a consequence, countries across the globe are engaging in efforts to speed up their domestic development of artificial intelligence (AI). In 2017, China showed its ambition to become a world leader in AI technology through the “New Generation Artificial Intelligence Plan'' (henceforth: 2017 AI Plan). Long-term industrial policies and innovation plans of this kind are by no means something new in China. However, what makes this plan especially interesting is that the key players in AI research and development (R&D) are actually mostly non-state-owned enterprises. Given these novel circumstances, which role does the state play in China’s AI development and how does it relate to the business actors in the field? This piece will briefly introduce the academic literature on the role of the state in Chinese industrial policy planning over the previous decades. It will then use a range of evidence from academia, media reports, and think tanks to show that new frameworks are needed to understand the relationship between business and state in China’s AI innovation in more recent years.
Since adopting the economic reform policy in 1978, industrial upgrading has always been at the core of China’s political and economic agenda. However, the methods adopted to achieve innovation have changed over the years. While most technology was acquired from foreign investors in the first period of economic reform, focus shifted towards domestic R&D in the 2000s, resulting in a more nationalist outlook of industrial policy (Naughton, 2018, pp. 344-56). Since then, Chinese R&D intensity (R&D expenditure as percentage of GDP) has increased steadily and is now on par with the EU average of around 2.1% (Eurostat 2020). Not only is the amount of investment in R&D noteworthy, but also its specific target: the so-called strategic emerging technologies. The rationale behind this is the belief that new technologies offer an opportunity for sudden, “leapfrog” catch-up with leading countries that would not be possible with conventional technology (Roberts et al., 2020, p. 5).
The extent to which the Chinese state directly controlled the development of these technologies depended on the strategic importance of the industry in question. Mattlin (2009) provides a framework based on three different categories of strategic relevance. Firstly, there are “strategic and key industries” that entail all industries with direct connection to national security. R&D in these industries would be executed by enterprises that are 100% state-owned and supervised by the “State-Owned Assets Supervision and Administration Commission” (SASAC). Secondly, there are “basic and pillar industries” that entail industries with implications for the wider economy, such as base metals or chemicals. State ownership in these industries would be reduced, but nevertheless substantial. All other industries belong to the third category. These “other industries” would generally be left to the private sector, although there would obviously still be some degree of surveillance by the state.
Mattlin (2009) himself considered IT and software to belong to the category of ‘Basic and Pillar Industries’. However, there is plenty of evidence that nowadays China sees IT and artificial intelligence as having key national security implications. For example, Allen (2019, p. 3) demonstrates that the CCP leadership unanimously considers AI to be crucial for great power competition. This is also directly reflected in the 2017 AI Plan which contains the term “national security” a total of eight times (Chinese State Council, 2017). The reason why AI plays such an important role in China’s grand strategy is again the idea of “leapfrog” development. It may be hard for China to catch up with the US army in terms of conventional military capabilities, yet, in areas like espionage or cyber-attacks, a sudden breakthrough might lead to a quick catch-up (Fritz, 2008). Apart from international competition, the Chinese state also sees AI as an important tool for securing social stability and surveillance at home (Roberts et al., 2020, p. 7). In Mattlin’s framework, this would mean that AI development in China should be executed by SOEs. However, in reality none of the cutting-edge information technology firms in China is state-owned (Naughton, 2018, p. 62). Hence, Mattlin’s theory is not very helpful in understanding the relationship between business and state in China’s contemporary AI development. As an alternative to that, I will propose two possible new frameworks.
The first is what I call “cooperative nudging”. It refers to the idea that business and state work together to advance China’s AI development by aligning each other’s incentives. As SOEs lack AI expertise and entrepreneurial skills, government officials who want to advance AI are incentivized to cooperate with non-state-owned companies (Yu, 2018, p. 6). To achieve this cooperation, the Chinese Ministry of Science and Technology nominated Baidu, Alibaba, Tencent, and iFlytek into a so-called “national team” just a few months after issuing the 2017 AI Plan (Jing and Dai, 2017). In 2018, Sensetime was added to the team (Ding, 2019). The idea behind the “national champions” is to create a win-win situation: the government gets to have a say in the macro-strategy of the company, while the company gains preferential political treatment (Roberts et al., 2020, p. 3). Ultimately, the goal is to make these national champions more competitive internationally as well (Hemphill and White, 2013). In the case of the AI national champions, this meant concretely that the government could determine a core research area for each of the companies: autonomous driving for Baidu, smart cities for Alibaba, medical imaging for Tencent, intelligent voice for iFlytek, and intelligent vision for Sensetime (Ding, 2019). In this way, the government does not need to have any ownership in the companies, but can simply “nudge” them to voluntarily work on areas the government wants.
In exchange, the national champions also receive substantial support from the government. For example, the government ensures planning security for the companies by providing significant funding and a favorable legal environment (Ding, 2018, p. 4). Another crucial mechanism through which the government tries to benefit its national champions is through technology standard setting. Even though we may rarely think about it, international standards are important for almost any technology we use in daily life. When purchasing a USB drive from one company in the US, and a laptop from another company in China, the USB drive will fit into the USB port of the laptop. That is because there is an international standard for USB. As most of the current technology standards come from outside China, Chinese technology companies still pay huge amounts of money on licensing fees every year (Koty, 2020). In response, the Chinese government announced the “China Standards 2035” plan in 2020. It aims to make most standards across different technologies, including AI, to be set by China by the year 2035. If successful, this would be yet another indirect measure with which the Chinese government can support its AI companies. La Bruyere and Picarsic (2020, p. 8) summarize this strategy by stating that “if you set the standards for how self-driving cars operate, you can make sure that your State champions – in self-driving cars and in their batteries – win, with or without subsidies.”
Despite all these efforts for cooperation between government and business in China’s AI development, some evidence also suggests that there is fierce competition for control between the two. I call this second framework an “emerging power struggle” between government and big business. A very recent example for this is the abrupt suspension of the planned IPO for Alibaba’s Ant Financial Group on November 2nd 2020, which many observers believe to be a signaling effort by the CCP to demonstrate who has the final say in China (Curran, Costa, and Chen, 2020). In the same month, the State Administration for Market Regulations also issued the “Anti-Monopoly Guidelines for the Platform Economy Sector” which directly target the very same companies that are also AI national champions (Baruzzi, 2020).
However, this is not to suggest that the state can simply do whatever it wants. A useful illustration of this is a dispute between China’s ride-hailing company, Didi, and the government (Vaswani, 2019). After two passengers had been killed by Didi drivers, the Chinese government tried to use this as a pretext to gain access to more of Didi’s consumer data. Due to the strong government pressure, Didi eventually had to comply. However, it also found its own creative way to fight back: the company handed over three large boxes of unorganized data printed on paper, making it essentially impossible for the authorities to process the data in any meaningful way. Sacks (2018) concludes from this case that there is “a kind of tug of war between the government and companies” in China.
In conclusion, the relationship between the Chinese state and big businesses in AI development is a complicated and subtle balancing act. On the one hand, the state adopts “cooperative nudging”: genuine efforts to cooperate by only indirectly incentivizing intended behavior of large AI companies, while also actively supporting them through the “AI national team” and international standard setting. On the other hand, there are also signs of an emerging power struggle between the state and large AI companies. For the time being, both of these mechanisms exist in parallel, and it is too early to judge which of them may eventually become the dominant feature of business-state relations in China’s AI development. The future trajectory of the relations will also be influenced by the extent to which both sides can create a clear position on their own. My analysis generally assumed both the state and a specific AI company to be a unitary actor with clear goals. In reality, however, both of them are already highly complex entities with multiple conflicting agendas on their own. It was beyond the scope of this essay to investigate if, for example, local and central government follow similar aims in AI industrial planning. It will also be interesting to observe which consequences the business-state relation of China’s AI development will have on global AI governance norms.
Gabriel Wagner is a German student in the International Studies BA program at Leiden University, where he specializes in the East Asian region and Mandarin language. Apart from his academic interest in Chinese politics, Gabriel also has a strong connection to the country through the mind sport Go. Before starting university, he spent 6 months studying Go in Beijing with a full scholarship from Tsinghua University. Now he is the manager of the youth Academy of the European Go Federation for which he organizes regular joint study activities with Chinese Go schools. Furthermore Gabriel is passionate about discovering evidence-based ways to solve important problems and is engaged in Positive Impact Society Erasmus (PISE) Rotterdam. You can follow him on LinkedIn.
The opinions expressed here are those of the writers and do not represent the views of European Guanxi.
Do you have an article you would like to share? Write for us.
Allen, G. C., 2019. “Understanding China’s AI Strategy: Clues to Chinese Strategic Thinking on Artificial Intelligence and National Security.” Center for a New American Security Washington, DC. Available from: https://www.cnas.org/publications/reports/understanding-chinas-ai-strategy [Accessed 20 Dec 2020].
Baruzzi, S., 2020. “China Releases Anti-Monopoly Guidelines for its Platform Economy.” China Briefing (December 16, 2020). Available from: https://www.china-briefing.com/news/china-releases-anti-monopoly-guidelines-for-its-platform-economy/ [Accessed March 29 2021].
China-State-Council, 2017. “China’s ‘New Generation Artificial Intelligence Development Plan’ (2017).” Translated by Graham Webster, Rogier Creemers, Paul Triolo, and Elsa Kania. DigiChina. Available from: https://d1y8sb8igg2f8e.cloudfront.net/documents/translation-fulltext-8.1.17.pdf [Accessed 14 Mar 2021].
Curran, E., Costa, S. H. e, Chen, L.Y., 2020. “Derailing of Jack Ma’s Ant IPO Shows Xi Jinping’s in Charge.” Bloomberg Quint (November 10, 2020). Available from: https://www.bloombergquint.com/global-economics/derailing-of-jack-ma-s-mega-ant-ipo-shows-xi-jinping-s-in-charge [Accessed 29 Mar 2021].
Ding, J., 2018. “Deciphering China’s AI dream.” Future of Humanity Institute Technical Report. Available from: http://www.fhi.ox.ac.uk/wp-content/uploads/Deciphering_Chinas_AI-Dream.pdf [Accesssed 15 Mar 2021].
Ding, J. 2019. “China’s AI “National Team”. ChinaAI (May 20, 2019). Available from: https://chinai.substack.com/p/chinai-51-chinas-ai-national-team [Accessed Mar 15 2021].
Eurostat, 2020. “R&D expenditure in the EU at 2.19% of GDP in 2019.” (November 27, 2020). Available from: https://ec.europa.eu/eurostat/web/products-eurostat-news/-/DDN-20201127-1#:~:text=In%202019%2C%20the%20Member%20States,compared%20with%202.18%25%20in%202018. [Accessed Mar 29 2021].
Fritz, J., 2008. “How China Will Use Cyber Warfare to Leapfrog in Military Competitiveness.” Culture Mandala 8 (1): 28–80.
Hemphill, T. A. and White, G. O., 2013. “China’s National Champions: The Evolution of a National Industrial Policy — or a New Era of Economic Protectionism?” Thunderbird International Business Review 55 (2): 193–212.
Jing, M. and Dai, S., 2017. “China Recruits Baidu, Alibaba and Tencent to AI ‘National Team’.” South China Morning Post (November 21, 2017). Available from: https://www.scmp.com/tech/china-tech/article/2120913/china-recruits-baidu-alibaba-andtencent-ai-national-team [Accessed 20 Dec 2021].
Koty, A. C., 2020. “What is the China Standards 2035 Plan and How Will it Impact Emerging Industries?” China Briefing, Available from: https://www.china-briefing.com/news/what-is-china-standards-2035-plan-how-will-it-impact-emerging-technologies-what-is-link-made-in-china-2025-goals/ [Accessed 15 Mar 2021].
La Bruyere, E. d. and Picarsic, N., 2020. “China Standards 2035: Beijing’s Platform Geopolitics and ‘Standardization Work in 2020’.” Horizon Advisory, China Standards Series, Available from: https://www.horizonadvisory.org/china-standards-2035-first-report [Accessed 15 Mar 2021].
Mattlin, M., 2009. “Chinese Strategic Sate-Owned Enterprises and Ownership Control.” BICCS Asia paper 4 (6): 1–28.
Naughton, B. J., 2018. The Chinese Economy: Adaptation and Growth. Cambridge and London: MIT Press.
Roberts, H., Cowls, J. , Morley, J. , Taddeo, M., Wang, V., and Floridi, L., 2020. “The Chinese Approach to Artificial Intelligence: An Analysis of Policy, Ethics, and Regulation.” AI & SOCIETY, 1–19.
Sacks, Samm. 2018. “What I Learned at Alibaba’s Data Protection Summit.” CSIS Commentary (October 11, 2018). Available from: https://www.csis.org/analysis/what-i-learned-alibabas-data-protection-summit [Accessed 15 Mar 2021].
Vaswani, Karishma. 2019. “Huawei: The Story of a Controversial Company.” BBC (March 6, 2019). Available from: https://www.bbc.co.uk/news/resources/idt-sh/Huawei [Accessed 15 Mar 2021].
Yu, Fu Lai Tony. 2018. “Private Enterprise Development in a One-Party Autocratic State: The Case of Alibaba Group in China’s E-Commerce.” Issues & Studies 54 (01): 1–33.