Hong Kong to focus on infrastructure monitoring in first phase of ‘AI city brain’

Hong Kong authorities will focus on incorporating infrastructure monitoring into the first phase of its AI-powered urban management system, while exploring its expansion into public services in later stages.

Officials also told a Legislative Council meeting on Tuesday that government arms were encouraged to develop their own databases before connecting them to the “AI city brain”.

Undersecretary for Innovation, Technology and Industry Lillian Cheong Man-lei said authorities would prioritise incorporating several databases monitoring city infrastructure conditions into the AI city brain system. These include the Lands Department’s geospatial database and its underground utilities system that maps water, natural gas pipes and electricity grid conditions.

The Buildings Department’s system which transforms two-dimensional building plans into 3D models, alongside information systems on highways, weather and landslide monitoring, would also be connected, she added.

“Our objective is to build a smart decision platform to collect, share and analyse interdepartmental data, with the aim of enhancing the government’s capability in resource allocation and decision-making, as well as prediction and alerts, ultimately improving our urban resilience,” she said.

The meeting was a joint session between the subcommittees on artificial intelligence (AI) and robotics development and “new quality productive forces” – referring to productivity led by technological advancement – to discuss industrial applications of these technologies.

The initiative stems from Chief Executive John Lee Ka-chiu’s latest policy address, in which he announced the government would develop the AI city brain system to consolidate databases maintained by different government arms and create an intra-government data-sharing environment.

A task force chaired by Chief Secretary for Administration Eric Chan Kwok-ki will steer the programme.

In an earlier interview with the South China Morning Post, Chan said the project aimed to provide a “full picture” of Hong Kong, and the authorities would launch a pilot version focusing on extreme weather response before typhoon season next year.

The first phase is expected to be completed in the first quarter of 2028, with HK$200 million (US$25.48 million) earmarked for this stage’s development.

During Tuesday’s meeting, several lawmakers asked whether the government would connect public service databases, such as the Labour and Welfare Bureau’s welfare recipient records, to the AI system to expand its functions beyond city management.

Acting Commissioner for Digital Policy Daniel Cheung Yee-wai said while the early stage of the project would focus on urban management and emergency response, authorities were in parallel exploring scenarios for later incorporation.

“We encourage various policy bureaus to work on databases within their respective purview and use AI platforms provided by the Digital Policy Office for analysis,” he said, adding that the Labour and Welfare Bureau has already been developing a system to identify vulnerable cases warranting more help.

“Ultimately, the AI city brain will connect these data sets for interdepartmental data analysis,” he added.

The meeting also discussed authorities’ plan to commission the government-owned Hong Kong Science and Technology Parks Corporation to develop the Hong Kong Innovation and Technology Application Scenario Promotion Centre.

The initiative, also announced in the latest policy address, aims to provide a targeted list of public and private sector scenarios where the tech industry can apply innovative solutions.

Cheong told lawmakers that another objective of the centre would be to facilitate trials of technological application and empower enterprises to better attract investment.

“In the Northern Metropolis … we expect to launch some new trial scenarios for new technologies, and the government will also provide policy packages such as subsidising public organisations to try these technologies,” she said.