On July 20, the Chinese government released its Next Generation Artificial Intelligence Development Plan.  The plan has gained significant media attention in part because it links AI with another topic that has drawn a considerable amount of attention, China’s “social credit system” (社会信用体系). Social credit uses big-data collection and analysis, to monitor, shape and rate individual’s behavior. While advances in artificial intelligence, and the growth of the surveillance state are all noteworthy on their own, China’s social credit program explicitly links them as parts of a broader political control process known as “social management” (社会管理).
The phrases “social management”, and the more recent version “social governance”, may seem like pseudo-scientific jargon, but in fact, are given clear importance by China’s top leaders.  In 2016, General Party Secretary Xi Jinping highlighted the concept, noting: “people working in political and legal affairs and comprehensive social governance have focused on dealing with outstanding problems and innovating social governance methods in recent years, achieving greater results,” (Xinhua, October 12, 2016). Elsewhere, the Party has clearly explained that it sees operationalizing social management as its blueprint for maintaining power. Far from being a narrow, isolated political concept, “social management” gives cohesion to an array of concepts ranging from Hu Jintao’s signature “Scientific Development” to Xi’s push for military-civil integration, as part of this power maintenance process (People.com.cn, April 17).
Social managements’ roots are in the core ideology of the Chinese Communist Party. The CCP defines itself as the “vanguard of the people”—the Leninist idea that a small group of scientifically guided and educated cadres can lead the people in the direction of social equality and prosperity. Mao Zedong’s organizational guide, the “mass line” describes the same concept. The CCP leadership is explicitly at the top of this hierarchical mass line system. It takes the “scattered and unsystematic ideas of the masses” and forms them into “concentrated and systematic ideas” before taking them back to the masses to “propagate and explain these ideas until the masses embrace them as their own”—Meaning management along scientific principles. 
Social management describes a “scientific” Leninist machinery for shaping, managing, and responding. It is best summarized as a complex systems management process through which the Party leadership attempts to manage the Party itself, and through which Party leadership attempts to manage the Party’s interactions with society as a whole. Social management is aimed at ensuring China’s “holistic” or “comprehensive” state security (国家安全). This holistic state security concept is not fundamentally new under Xi Jinping. It includes the western “national security” concept, but, more significantly, is focused on two internal security dimensions. First, managing the Party itself, and second is managing social order (Xinhua, April 15, 2014; Qiushi, April 15, 2017; PLA Daily [archive], December 13, 2000).
Social management itself is not a new concept and dates to the PRC’s founding in 1949, when it was first integrated into the CCP’s discussion of law and social order. The concept became increasingly prominent in the Party leadership’s rhetoric between the late 1990s and early 2000s. When the 12th Five-Year Plan for National and Social Development was released in March 2011, social management was enshrined as a key objective (China.com.cn, March 16, 2011). In the plan, the Party set targets for speeding up the construction of a social management system that combined governance measures to address problems at their source, dynamic management, and emergency response—while adhering to the core leadership of the Party.
In its ideal form, the social management process optimises interactions vertically (within the Party), horizontally (between agencies), and holistically, between the Party and society. At every linkage, the goal is to improve governance capacity to shape, manage, and respond to social demands. Social management must efficiently solve problems to succeed. Such problems include: allocation of public resources, preventing and controlling risks associated with man-made and natural disasters, stopping dissent, and pre-empting and managing social conflict. The process involves both coercive and co-optative tactics, constantly acting together, to force individuals and to incentivize individuals to participate in social management.
For the social management process to succeed, particularly when in a crisis response mode, an automation of the interactions between the state and society, as well as the interactions within the Party itself, is required. The modern-day “grid management” (网格化管理) and the “social credit system” (社会信用体系) are unique compared to previous versions of similar social control mechanisms because they employ modern technology. They represent the attempted automation of social management.
The concept of automating social management is not new under Xi Jinping (2012-Present) or his predecessor Hu Jintao (2002-2012). In fact, the concept emerged in the late 1970s when “social management” was directly connected to complex systems theories the Party-State’s theorists were drawing from to design a Leninist governance system to recover power after Mao and the Cultural Revolution. The basic ideas originated around 1957 when Qian Xuesen (“father of Chinese rocketry”) called on the Chinese Academy of Social Sciences to take the concept seriously as a way of solving social problems [People’s Daily [archive], 28 May 1957]. By the 1970s and throughout the 1980s complex systems thinking [largely via Qian Xuesen’s promotion of engineering cybernetics] was clearly tied to “social management”. For example, one report from 1984, “On the New Technological Revolution” (新技术革命) said:
Leaps and bounds in science and technology [since the 1940s have] influenced or given rise to transformations in the way social management agencies work. The theory and practice, perspective and method of systems engineering were born and developed from these changes.” It elaborated that it is impossible to manage effectively through individuals or a small number of people, and, “only if we fully grasp [the concepts of] information, data, systems analysis, and decision modeling, can we truly possess ‘foresight and sagacity’, and generate the courage and a bold vision consistent with the flow of history. 
The report further laid out what steps were needed to implement systems engineering in the “social domain”. It included, among other things, defining what targets systems management should reach, establishing facilities to ensure information flow, and planning and developing methods and procedures for systems analysis. This explains why systems thinking is key to understanding not only how social credit fits into social management, but overall how the social management system is being designed.
Rather than being relatively new conceptions, modern surveillance techniques and social credit are merely the newest developments in realizing the automated social management objective. Advances in AI and big data management further improve their function, from a technical perspective. These advances describe what the Party refers to social management “innovation”.
The first major step in the technological development of social management’s automation was the implementation of grid management (网格化管理). Structurally, it advanced what has been described as a multilateral “vertical and horizontal integration” (纵横结合) of resources, people and agencies involved in social management. The political-legal and public security apparatus, including neighborhood and street-level committees, largely responsible for the technical side of its day-to-day implementation. Grid management enabled the organization of data to generate better situational awareness and predictive capacity, as well as enhanced tracking and monitoring of individuals (People’s Daily, October 15, 2006).
The first modern grid(-ized) (网格化) policing was implemented between 2001 and 2002 in cities like Shanghai (People’s Daily [archive], August 3, 2001; China File, August 10, 2016). It organizes information gathering by dividing an urban space into grids, each of these grid spaces is assigned grid managers who help to collect data and pre-empt and solve problems within their grid. The modern informatized grid enables faster emergency response and improved prevention and control. The photographs and videos police take at the scene of almost every protest are one example of the kind of data fed into the grid system.
Grid management’s application to social management was significantly expanded between 2002–2012 under the direction of Zhou Yongkang, first as minister of Public Security and later as head of the Central Political-Legal Affairs Committee. Advances in integrated e-government resources in the internal security apparatus, namely the Golden Shield Project, greatly enabled grid management. The Golden Shield Project is not an internet monitoring project updating the Great Firewall. Rather it is an e-government project creating an organizational network connecting the Ministry of Public Security with its local-level bureaus, which was already being employed at provincial and city levels by 2002 (China Brief, June 3, 2011; People’s Daily [Archive], April 26, 2002). The “Shield” was part of an expanded series of systems engineering projects, originally initiated in 1993 and later expanded as “Golden Projects”. Each of the Golden Projects were e-government projects designed to build and streamline information systems, and connect agencies to improve their operational capacity.
This eventually included the multi-phase Golden Shield project, which was being implemented under the guidance of the State Informatization Leading Small Group by the late 1990s and early 2000s.  For public security bureaus, it improved both efficiency and surveillance. Software applications were developed to integrate data by requiring “real name” registration for travel booking, telecom services, and other services, information from hotel check-in and at customs clearance could be linked to law enforcement databases. The major contribution of the Golden Shield Project to the overall social management program was that it created a capacity to automate information sharing. Ostensibly, the Golden Projects were the technological starting point for building the social credit system, and perhaps social credit was an end goal much earlier in the process. E-government in China has always been designed to improve governance capacity and operate as a feedback loop with social management functions. The timing of social credit implementation probably is explained more by improved technical capacity than by changing policy objectives.
Automated Social Management?
The social credit system relies on the technology enabling and the organizational capacity created through the grid management system. Effectively “social credit” is the technological marriage of individual “responsibility” mechanisms and social control methodologies. Responsibility is a concept underlying the social management process, and it implies that the entire Party and all of society are responsible for upholding the Communist Party’s leadership. This is also why individual responsibility is a key theme of all major state security-relevant legislation passed under Xi Jinping (IISS Voices, May 26; The National Interest, May 17, 2016). Enabled through the same resources and technology found in grid management, social credit creates a simultaneous co-optative and coercive responsibility systems function, and when fully implemented comprehensively covers all of society. Society is co-opted to participate because the same technology is directly linked to conveniences that improve everyday life, for instance electronic payment. Society is coerced to participate, for instance by self-censoring online, because increasingly technology systems are improving the government’s capacity to enforce “responsibility” to the party-state. Not participating could have consequences not only for the individual but also their personal networks. These functions will only become further advanced through plans such as “Internet Plus”, as the same technology applications used to provide social and commercial services feed directly into government information gathering and sharing processes (Gov.cn, February 1).
In the construction of the social credit system, current research and development is largely focused on areas such as big data analysis and integration to support the collection of information and ensure its effective use for intelligence. This is one of many areas where advances in artificial intelligence would help streamline social management processes and, perhaps ideally, even automate them. Two major problems, however, confront this automated version of social management.
The first is the struggle for power within the Party. The Party members in charge of day-to-day implementation of social management are also responsible to the Party. As the systems were being enabled in the early 2000s, these agencies had a large amount of relatively unregulated power. The age-old problem of an authoritarian system is that security services require substantial power in order to secure the leadership’s authority. The same resources enabling management of the Party-society relationship can be abused by Party members and used against other within the Party (War on the Rocks, July 18, 2016). This appears to be the case with Zhou Yongkang, Bo Xilai, and others ahead of the 18th Party Congress. The problem will not disappear in a Leninist system, which not subject to external checks and balances. And it is why ensuring loyalty is a major part of the management of the party side of “state security”.
The second probably is a symptom of the first: disaggregated security agencies. In an ideal form, agencies tasked with different aspects of social management can cooperate to address state security problems that have “integrated” characteristics. Usually such threats involve the ‘three evil forces’ of splittism, terrorism and extremism, and often specifically are related to Tibet, Xinjiang and Falun Gong. Because these are described as threats that have domestic and international connectivity, cooperation between domestic departments, intelligence, and foreign affairs is required for operational success . It is particularly applicable in massive multi-agency operations such as “Operation Skynet”, tracking down fugitives from the Party-state (SCMP, March 26, 2015).
Both problems are explanations for structural changes that put Xi Jinping in charge of leading groups on State Security, Cyber Security, and so on. Using the example of the Central State Security Commission, there are now local government-level iterations in the form of state security work small leading groups in nearly every province, as well as the counties and cities within them. All are led by the relevant party secretary of the locality, and, where data is available, their membership appears to include (but is not limited to) the heads of Political-Legal Affairs Committees, Ministry of State Security bureaus, Armed Police, and Propaganda departments. Similar committees have been set up to mirror other new central leading groups. The membership overlaps significantly. Such leading groups are not new, but the evidence points to the system being utilized not only to re-center power away from the Central Political-Legal Affairs Committee (CPLC) and local versions, but also to develop a more effective system for mobilizing the social management process. For as much as the changes may be geared toward re-centering internal security power, the changes probably serve a dual purpose of creating a capacity for departments to function like a holistic “system of systems”. It would address problems by issue—rather than as separate systems addressing overlapping problems.
Chinese information technology research and development, including the priorities outlined in the artificial intelligence plan, are interesting on their own because they mark advances in important research areas. But, as the language of the AI development plan indicates, these advances cannot be separated from Beijing’s social management and state security policy. Applied to the social management process, they are aimed at improving governance capacity—automating the “carrot” and “stick” processes that ensure the Party-state’s power. Senior CCP leadership hopes that through automation the Party will be able to more effectively anticipate and react to emerging problems, pre-empting a crisis before they become serious threats to stability.
Samantha Hoffman is an independent consultant. She is currently finalizing her Ph.D.: “Programming China: The Communist Party’s Autonomic Approach to Managing State Security”. She tweets @he_shumei
- A translation of the plan, completed by Graham Webster, Paul Triolo, Elsa Kania and Rogier Creemers, is available at China Copyright and Media blog: https://chinacopyrightandmedia.wordpress.com/2017/07/20/a-next-generation-artificial-intelligence-development-plan/
- Social management (社会管理) and social governance (社会治理) are two phrases that, in practice, have the same definition and are implementing exactly the same process, but the shift from “social management” to “social governance” under Xi Jinping has more to do with political power and ensuring the effectiveness of the social management process, than actual conceptual change. Hoffman, Samantha. “Portents of Change in China’s Social Management.” China Brief 12, no. 15 (2012): pp. 5-8; Hoffman, Samantha, and Peter Mattis. “China’s Proposed “State Security Council”: Social Governance under Xi Jinping,” China Policy Institute: Analysis, 21 November 2013. accessed via China Policy Institute.
- Heath, Timothy. “Xi’s Mass Line Campaign: Realigning Party Politics to New Realities.” China Brief 13, no. 16 (2013): 3-6. Mao, Zedong. “Some Questions Concerning Methods of Leadership ” Marxists.org, 1 June 1943, https://www.marxists.org/reference/archive/mao/selected-works/volume-3/mswv3_13.htm.
- The People’s Daily [archive], September 13, 1984
- Zhou, Hongren, Hongyuan Xu, Yuxian Zhang, Changsheng Wang, and Xinhong Zhang. China E-Government Development Report No. 1 (中国电子政务发展报告) Blue Book of Electronic Development (电子政务蓝皮书). 2003.
- UIR Center for International Strategy and Security Studies (中国关系学院国际战略与安全研究中心), Annual Report on China’s National Security Studies 2014 (中国国家安全研究报告 2014), Blue Book of National Security (Beijing: Social Sciences Academic Press (China), 2014), p.133.