The collection begins with Jeffrey Ding’s essay on China’s AI innovation ecosystem. Contrary to western perceptions that China’s AI development is driven by a centrally planned industrial strategy, it is at the local level that real innovation is happening. Cities are creating successful ‘hybridised industrial ecosystems’ by connecting the public sector, leading tech companies, research institutes and networks of SMEs to drive technological advances and implementation. Increased R&D spending is only part of the equation: for such ecosystems to develop successfully, cities also need a strong local technology base, an elite university, a co-operative local government, a good skill supply and inter-regional or international linkages.

Andy Chun is optimistic about AI’s potentially revolutionary impact on healthcare in China, which faces the dual challenges of an ageing population and a scarcity of medical services in rural areas. AI is being developed and deployed in several key areas: improving diagnostic accuracy; assisting with treatment decisions; predicting potential health problems through analysis of patient data generated by wearables; automating certain tasks to reduce physician workload and increase remote access; and reducing R&D timeframes for new pharmaceuticals using AI modelling and simulations. Following the outbreak of COVID-19 in Wuhan, these key AI applications are being deployed in China to tackle the pandemic and contribute to vaccine research efforts. China’s ‘experiment first, regulate later’ approach is aiding its rapid advances in healthcare AI - in a European context, regulatory innovations such as sandboxes could help boost the pace of progress.

Yi-Ling Liu examines the emergence of AI educational tools and adaptive learning systems, and the potential benefits that these innovations are said to offer. However, with development led in large part by private sector edtech startups, it remains to be seen whether these tools can translate to the public education system and help balance the rural-urban educational divide, as touted, or whether in fact they are likely to entrench existing socioeconomic inequalities. Moreover, there are fundamental questions around whether AI educational tools are beneficial in a pedagogical sense, capable of teaching students beyond mere rote learning, as well as serious concerns about the ethical implications of collecting student data. Learning from China’s experiences of AI in education becomes of vital importance as the world looks to online learning as a potential solution to the global disruption to schools due to the coronavirus pandemic.

Eugeniu Han examines the use of AI in smart cities in China, exploring two illuminating case studies: ‘City Brain’, an intelligent urban management system designed in Hangzhou, which has now been rolled out to 23 cities across Asia, and Shanghai’s Smart Court system. While the judicial system is not often analysed alongside urban management and transport services in typical smart city case studies, this original perspective reflects two key drivers behind the development of China’s smart cities: the growing demand for improved services in the context of urbanisation, and the desire to strengthen China’s main governing institutions - this latter theme is also explored in the essays by Dev Lewis and Rogier Creemers. The topic of smart cities in China is complex: some applications have great potential to benefit cities and citizens and make urban services more efficient, yet other elements are problematic, especially regarding issues of privacy, bias and political interference.

Dev Lewis examines the development of China’s Social Credit System (SCS). The SCS has gained notoriety outside China and, while critics raise legitimate concerns, there is much misunderstanding about the system’s functioning, aims and reach. The SCS forms part of the Chinese government’s attempt to strengthen its weak institutions and boost trust between governments, firms and individuals by improving and harnessing its public data platforms. Rather than being a single overarching system, the SCS is a cluster of different emerging initiatives. These include the national Blacklist-Redlist Joint Sanctions and Rewards regimes that attempt to remedy the judicial system’s failure to enforce laws, and municipal-level Citizen Score experiments being developed in some cities that offer benefits to individuals with ‘good credit’. Western narratives tend to conflate SCS with other types of surveillance in China, obscuring what is distinctive about the system and the motivations behind it. Of course, there are causes for concern too - particularly over fairness, accountability and the lack of transparency around what is deemed to constitute ‘trustworthy’ behaviour.

Danit Gal explores China’s approach to AI ethics, dispelling a common assumption that these issues are not being considered in China. China has established various expert committees and published several documents outlining AI ethics principles. In some areas, especially around privacy, China’s approach aligns with universal ethical principles adopted globally, and the international community could do more to include China in global AI governance discussions. But China also has its own unique ethical and cultural foundations, and questions around how to apply these distinct philosophical concepts to AI are still being debated in China. Applying AI in public services, as explored in the collection’s other essays, carries with it a host of ethical considerations and implications, including privacy, bias, accessibility, accountability and transparency. While China has implemented robust consumer privacy laws, the government’s use of private data for surveillance is a cause for serious concern. Yet it is worth remembering that governments in Europe and the US are also using facial recognition in problematic and concerning ways. The scope and extent of China’s deployment of AI in public services may be beyond anything occurring elsewhere, but the ethical considerations it raises are universally relevant, and examining China’s experience helps us reflect on shared ethical issues that must be tackled collectively.

Finally, Rogier Creemers offers a critical reflection on why China is seeking to deploy AI in public services. China’s drive to deploy technology in social governance stems from the Communist Party’s ideological view that social order is governed by an objective and intelligible set of ‘laws’, and that big data and AI can both help to understand these laws and ‘engineer’ society to solve social problems. The development of this ideological position is particular to Chinese intellectual history, yet there is an interesting parallel to be drawn between this ideological position and Silicon Valley ‘solutionism’, where technology and data are seen as a panacea for social problems. Strikingly, China’s State Council explicitly identifies AI as a tool to assist with both sides of ‘social governance’: providing better public services as well as preventing and controlling social unrest. Understanding this helps us see that the smart courts in Eugeniu Han’s essay and the Social Credit System in Dev Lewis’ essay are part of the same story as the other public service AI applications that are intended to improve the lives of citizens.