BY Zhu Rongsheng & Chen Qi

With China entering a new development era, technological innovation and social governance play an increasingly central role in the country's overall modernization. The Chinese Government has put forward a "three-step" strategy for AI governance, aiming to become a world player in terms of AI theory, technology and application by 2030 and establish more complete AI laws, regulations, ethics and policy systems.

To achieve the development goals, the Chinese Government is striving to increase the number of pilot zones for next-generation AI innovation and development, thus speeding up the transformation and application of research findings. AI cybersecurity system is also a pressing concern. The policy deployment reflects China's efforts to address the potential contradictions between technology development and practical application.

So, how do we understand China's governance model from an evolutionary perspective?

In February 2019, China's National Governance Committee for the New Generation Artificial Intelligence proposed eight principles.

The AI Governance Principles provide a framework and action guidelines for AI governance, aiming to "promote the healthy development of a new generation of AI; better coordinate the relationship between development and governance; ensure that AI is secure, reliable and controllable; promote economically, socially, and ecologically sustainable development; and jointly build a community of common destiny for humanity."

The eight principles of AI governance are proposed as harmony and friendliness, fairness and justice, inclusiveness and sharing, respect for privacy, security and controllability, shared responsibility, open cooperation and agile governance.

According to Xue Lan, a professor at Tsinghua University's Institute for AI International Governance, compared with the passivity of the previous responsive governance model and the high cost of the centralized governance model, the core of agile governance is to recognize and solve problems in a timely fashion, to avoid the AI technology becoming something "evil". Fu Ying, who is now chair of the Center for International Strategey and Security at Tsinghua Universty, former Chinese vice-foreign minister, noted that traditional legislation process takes a longer time to reach social consensus and international rules, but the iterative development of AI technology means that there may not be the luxury of time to address the challenges posed by AI applications. The agile governance model provides a feasible way for various stakeholders to participate in improving governance efficiency.

Confronted with the uncertainty, complexity and unknown of cutting-edge technologies, the innovative agile governance model provides a feasible way for all stakeholders concerned with AI development to improve the efficiency of governance.

Agile governance is a law-based, rule-led, market-driven measure to avoid rigid management hindering the development of technology and the growth of social welfare. Agile governance emphasizes the establishment of an ecosystem in which innovation subjects, and governance subjects, respond to technological changes in a timely manner, thus providing an inclusive and sustainable innovation environment for sustainable technological development.

China is stepping up the construction of legal norms and guiding rules to enable multiple actors to jointly use AI technology for the sake of social welfare.

Electronic Commerce Law, Civil Code, Data Security Law and other laws stipulate that governments, enterprises, and individuals shall not collect information of citizens excessively, and cannot divulge, sell, or illegally provide personal information to others.

Since 2018, the Standardization Administration of China and China Electronic Standardization Institute have published white papers on AI standardization every year to promote the technological development path of privacy protection, and cope with the challenges of commercialization to ethics and laws from the very beginning of research and development. Chinese leading AI enterprises have set up special governance institutions, developed application guidelines and technical tools for enterprise internal control, and participated in the process of formulating social guidelines for the joint application of technology.

For example, compared with laws, industry standards are more flexible, non-mandatory and autonomous. They not only complement "hard laws," but also can be more easily applied to the spontaneous needs of the international market. However, "soft laws" are confined by insufficient enforcement and strong reliance on stakeholders. The agile governance model needs to consider incorporating appropriate incentives into regulatory mechanisms.

The agile governance model is a result of China's intrinsic need for high-quality development, while drawing on experience with dealing with the complex challenges of external technological decoupling. This model provides a "Chinese solution" to the global governance path. At the present time of fierce debates about how to manage technological challenges, what we need is a sense of inclusiveness of different solutions, and walk towards the future of "science and technology for humankind" guided by a sense of human community.

Zhu Rongsheng, research associate at the Center for International Security and Strategy of Tsinghua University

Chen Qi, chief expert and professor at the Institute for AI International Governance of Tsinghua University