Data handling concerns and a lack of community engagement are addressed in different ways, with Beijing district a model, analysts said
Each night in the darkest hours, a fleet of drones patrols the eastern Chinese city of Nanjing, watching over potentially dodgy areas such as underground station exits and the riverside to deter any would-be criminals.
The nightly routine has been operating since last year, according to a report published in April by the city’s police bureau, which detailed its use of artificial intelligence to plan patrol routes.
According to the report, the AI puts a laser focus on patrolling Nanjing’s blind spots – the areas generally ignored by human patrols.
The city’s experience is part of China’s efforts to scale up the use of AI in urban management, to better handle complex challenges brought by the massive scale of city life and rapid urbanisation.
The Chinese leadership has long viewed AI as a tool for transforming the economy and refining its governance, launching the "smart cities" initiative and other programmes over the past decade.
Despite this rapid progress, analysts warned that China’s AI governance faces challenges – including privacy protection and limited community involvement – that are mirrored in other parts of the world as governments race to adopt the technology. The notion of intelligent governance, exemplified by the “"city brain” – an AI-powered central nervous system that collects and processes information that informs decision-making – officially became a Chinese national strategy in 2017.
Beijing, Shanghai, and the futuristic Xiongan New Area, 100km (62 miles) south of the capital, have developed their own city brains, according to various government reports. But detailed information on the exact progress and how they are being used is scarce.
State news agency Xinhua said in 2020 that more than 500 Chinese cities were in the process of building their own versions.
According to Xu Ying, an assistant professor in the Department of Urban Governance at the Central University of Finance and Economics, the city brain operating in Beijing’s Haidian district is seen as one of the best examples in China.
The district has been using AI to manage many aspects of public life, such as security, regulating traffic and dealing with the dangers of flooding.
But the project has faced the same concerns – including protecting data privacy and limited community involvement – that led to the cancellation of a similar project in Toronto in 2020, he said.
Xu, whose work has included in-depth field research into Haidian district’s city brain, said that the operation centre had done a lot to remove personal information from its data collection.
However, concerns remained that citizens’ personal information or image rights could still be exposed, he said.
“Many of the images and videos collected by the system capture people’s faces and this raises the question: can individuals’ personal movements or behavioural patterns be identified and potentially misused by certain departments or individuals?”
More than 12,000 surveillance cameras and over 10,000 sensors cover the district, obtaining 24-hour real-time information. Around 60 million data entries are collected every day, according to a report released by the Haidian district government in 2021.
The report said that the city brain could improve the efficiency of government officials and meant that “whatever they want to see is available at a glance”.
In Toronto, a similar attempt to build a futuristic community, dubbed the Quayside project, was initiated in 2017 by Sidewalk Labs, part of Google’s parent company Alphabet.
Sidewalk Labs aimed to incorporate smart technology, including the use of cameras and sensors that track people’s movements, to improve mobility and reduce traffic on Toronto’s lakeshore.
Instead, the project was abandoned, in part because of mounting concerns about the way data was handled and its lack of public engagement – a “negative example” that China “should take as a lesson”, according to Xu.
Canadian privacy law requires companies to gain consent before collecting personal information that relates to identifiable individuals. Sidewalk Labs argued that it would be difficult to gain people’s consent for data collection in public spaces.
A 2019 survey by Toronto-based market research firm The Forum Poll showed that only 38 per cent of the city’s residents were familiar with the smart city project. Of those, 60 per cent did not trust Sidewalk Labs’ collection of data.
Xu said that the development of Haidian’s city brain lacked public involvement in a similar way to the Toronto project, citing the Beijing district’s attempts to use AI to improve flood management as an example.
As part of the city brain project, Haidian district released an app around 2021 to sync urban flooding data and allow residents to track and avoid dangerous areas during the rainy season.
The city brain collected weather data and produced reports showing which spots were most prone to flooding. Officials then assigned resources to the areas in the most urgent need of attention.
The urban flooding data was also shared to residents via the app, which was called the Haidian Water Brain. “I interviewed some residents and found that very few were aware of the app, let alone had used it,” Xu said.
The app became unavailable to the public without explanation, according to Xu. He heard from insiders that the closure might be due to sensitivities around weather data, and potential national security risks in making it public, he said.
“The Haidian city brain project had very limited public participation throughout its planning, design, construction, and implementation phases. That’s one of the reasons ordinary residents don’t feel very connected to it or even aware of its existence.”
Xu said it would have been better if there was a transparency mechanism for public inputs. “Urban governance is fundamentally about interacting with people and society – it shouldn’t be just one side launching something and then considering things are done.”
The Haidian city brain’s unclear administrative ranking in the Chinese government also made it difficult to coordinate across departments – a factor that may eventually undermine its effectiveness, according to Xu.
“It would lead to questions – on what basis could it coordinate? Does it hold a higher administrative rank compared to the other departments? If not, its ability to influence other government agencies is limited,” he said.
Richard Hu, an urban planner and professor at the Canberra Business School, said China’s experience of advocating smart city and innovation-led urban management had given it a strong advantage compared with other countries.
“The disadvantages are that China is vast and has numerous cities that are unbalanced in development, including the hard capacity of technology and soft capacity in using tools like AI,” he said.
According to peer-reviewed research published by the Journal of Urban Technology in January 2024, the top 21 cities that most frequently raised AI in policy discussions were close to the coast an in the country’s traditional economic heartlands, with Beijing, Shanghai, and Shenzhen leading the list.
Only two cities from other parts of the country – Wuhan in central China and Chongqing in the southwest – were included in the ranking, the report noted.
“China’s approach is government-led and followed by industry and the community’s involvement is limited,” Hu said.
“Whether other countries can or should learn from China depends on many factors. The strong role played by the government and the top-down approach may not work in other countries,” he said.
Federico Cugurullo, an associate professor in smart and sustainable urbanism at Trinity College Dublin, said that while China was leading the way, the drive to apply large-scale AI to urban governance and make cities more intelligent was a global phenomenon.
“It is something that every government would like to have, because it is a very powerful way to control the city, to manage the city, but also to develop specific agendas, though this depends on the local context,” he said.
For example, the US Department of Homeland Security has been using AI tools since Donald Trump’s first term in office. These include facial recognition scanners in public areas to track down alleged illegal immigrants and robotic dogs patrolling its southern border.
According to Cugurullo, many of the AI-assisted urban scenarios in Europe were similar to China but on a much smaller scale, because of the more limited powers of European local governments.
Haidian’s city brain is run by the state-owned Zhongke Brain. The project involves the cooperation of more than 70 Chinese firms, including leading hi-tech company Baidu and telecommunication companies.
In contrast, when Barcelona city council in Spain wanted to use artificial intelligence to detect traffic infringements and improve bus services, it collaborated with a single US tech firm, Hayden AI.
According to a report by the Barcelona council, cameras were installed at the front of four buses in the pilot project. These automatically detected cars that were illegally parked in bus lanes or at stops as the buses travelled along their routes.
“There is this strong cooperation between the tech sector and the governments at national scale, at the local scale, and this is why China is moving forward very quickly … while a lot of our local governments don’t have much power,” Cugurullo said.
“They don’t have a lot of resources.” he added. According to Cugurullo, a project like Haidian’s city brain could not be developed in a European city because of the different capabilities, cultures and regulations.
---
Autor(en)/Author(s): Sylvie Zhuang
Quelle/Source: South China Morning Post, 31.05.2025
Bitte besuchen Sie/Please visit:
