The first one is the Beijing AI Principles (BAIP), released in May 2019, backed by the Ministry of Science and Technology and drafted by the Beijing Academy of AI in a multi-stakeholder consultative process. While the contents of the BAIP closely align with existing international principles, it has some distinctive features. The BAIP takes an applied approach to AI ethics by suggesting detailed and action-oriented goals, norms and best practices in research and development, AI use and AI governance. In addition, the document references the ‘philosophy of “optimising symbiosis”’. One of the drafters, Yi Zeng, explains that optimising symbiosis refers to the principle of ‘harmony’ in Chinese philosophy, where all life forms evolve in harmony rather than competition and is especially relevant in the context of human-machine interactions. This particular application of Chinese ethics and philosophy to AI is a distinctive approach that I will return to in this essay.

Finally, in line with its applied focus, the BAIP touches on long-term planning for the expected development trajectory of AI. The BAIP’s forward-looking, hands-on approach and long-term outlook is unique among other global principles, as is the way it links existing principles to Chinese practices and philosophical traditions.

The second document is a draft Joint Pledge on AI Industry Self-Discipline, also released in May 2019. It was backed by the Ministry of Industry and Information Technology and drafted by the China Academy of Information and Communication Technology in another multi-stakeholder consultative process. Again, this document largely aligns with other global AI ethics principles but has one notable point of distinction. Article 10: Self-discipline and self-governance, is unique in the global landscape as well as the Chinese one.

This article calls to ‘strengthen awareness of corporate social responsibility, integrate ethical principles into all aspects of AI-related activities and implement ethical reviews. Promote industry self-governance, formulate norms of behaviour for practitioners, and progressively build and strengthen industry supervision mechanisms’. The suggestion of ethical reviews, self-governance and industry supervision mechanisms introduces a significantly stronger degree of implementation, enforcement and oversight on AI ethics compliance, a field still largely considered soft law. This lends a more applied nature to AI ethics principles in the Chinese AI industry, in the event that this joint pledge materialises.

Finally, the third document is the Governance Principles for a New Generation of AI, released in June 2019, also backed by the Ministry of Science and Technology and drafted by the National New Generation AI Governance Expert Committee. As well as being aligned with other international AI ethics principles, these principles also share considerable similarity with the two previous Chinese documents. Predominantly applied in nature, this document again references harmony, offering a more direct translation as ‘human-machine harmony’. In a global landscape of human-centred AI ethics, this vision is unique in its holistic approach to the interactions between humans and AI, and the way they shape and are being shaped by each other.

An additional official committee still in discussion is the Professional Committee on AI Ethics under the Chinese Association for Artificial Intelligence, a state-level AI organisation under the Chinese Ministry of Civil Affairs. In a media interview, the committee’s leader, Professor Chen Xiaoping, shared his goal of developing ‘an entire system of AI ethics, with ethical guidelines and a set of operating mechanisms to realise them’.

These official committees are joined by ongoing corporate efforts to align with the developing AI ethics discussions. These efforts received considerable attention during the Chinese Government’s Two Sessions in March 2019, when the CEOs of technology companies Baidu and Tencent submitted proposals calling for AI ethics principles and rules. Baidu’s proposal called for the sharing of distinct Chinese wisdom with the international AI ethics community, and Tencent’s proposal focused on the societal applications and implications of AI.