Accountability in Algorithm-assisted Sanctions: Public Law Scrutiny of China’s Social Credit System
Principal Investigator: Dr. Clement Chen
Project name in Chinese:「借助演算法確定制裁」如何問責——從公法的角度檢視社會信用體系
Funded by General Research Fund (GRF). HKD 705,920 awarded in year 2020-2021 for 36 months.
About the project: The proposed project will investigate the accountability issues of the core mechanism of China’s Social Credit System (SCS), i.e. algorithm-assisted sanctions, and review ways to minimise the mechanism’s impact on citizens’ rights. The SCS was introduced in mainland China in 2014, and is expected to operate nationwide by the end of 2020. It is a far-reaching system of social and economic governance that harvests the power of datafication and algorithms. Concerns have been raised over the possibility that the SCS will become a social engineering project of unprecedented scale, affecting the lives and rights of over a billion citizens. This project constitutes a timely inquiry into the evolution and accountability of the system, and thus is of profound theoretical and practical significance. It will produce a comprehensive and critical account of an essential mechanism of the system, i.e. algorithm-assisted sanctions. That mechanism consists of two elements: profiling and sanction. Profiling refers to the rating of the ‘trustworthiness’ of individual citizens based on the datafication and analysis of their compliance with various state-endorsed rules, including but not limited to legal norms and ethics. Different punishments are then determined and imposed concurrently by various authorities according to the rating results as sanctions against untrustworthiness. Often called ‘combined punishment’ in policy documents, the algorithm-assisted sanction mechanism plays a pivotal role in the SCS. My project, which will be carried out in three work packages, will situate the sanction mechanism and its technological support in the context of administrative law. First, using the theoretical framework of algorithmic regulation and carrying out empirical investigation, it will outline and categorise the roles played by algorithms in the profiling and sanction of individuals. It will then examine the compatibility of different types of algorithm-assisted sanction with administrative law principles concerning the legality and reasonableness of decision-making. Second, based on a sizeable and representative sample of cases, it will review the adequacy of judicial remedies offered to individuals affected by algorithm-assisted sanctions. Finally, it will offer proposals for reforming doctrines and judicial review approaches to better safeguard individual rights under the SCS, on the one hand, and identify the structural issues of the SCS that need to be addressed by legislation, on the other. The key project output will include research articles on law & technology studies and doctrinal public law and submissions in response to public consultations on SCS-related legislation at the central and local levels.