From insight to action: What does it take to build an AI city?

Over the past decade, governments around the world have increasingly turned to smart city initiatives to improve the quality, efficiency and sustainability of city services, using digitalised sensors to collect and connect data across an area’s transport networks, pedestrian activity patterns, water consumption, waste levels, air quality, and more. But achieving these important objectives requires action, not just insight. As we enter the next stage of innovation, how can governments advance their cities from digitalisation to intelligence? From there, how can they turn predictions or requests into timelier services and rapid on-the-ground responses? Discover B2B Marketing That Performs Combine business intelligence and editorial excellence to reach engaged professionals across 36 leading media platforms. The answers are found in the ‘AI city’ vision, where digital foundations and data are layered with AI agents that detect patterns, forecast demands, recommend actions, and automate tasks, making services more accessible, responsive and efficient for all. A proactive approach to urban challenges Several real-world examples illustrated the AI city potential at the Global Government AI CITY Innovation Summit, held during Huawei Connect 2026 in Shanghai, China. The summit brought together more than 500 participants to highlight the rapid advancements that AI is now making in urban development. In his keynote speech, Simon Zhou, VP, Government Public Services Digitalization BU, Huawei, presented AI as a way for cities to respond more proactively to the challenges of growing populations and increasingly complex urban needs. In one notable example, he explained how an AI agent called Long Xiao Er (Longgang’s AI-native urban governance agent) played a critical role in helping emergency services in Shenzhen, China prepare for Typhoon Noul last July. Bringing together weather, sensor, and public report data to monitor vulnerable low-lying areas of the coastal city, the AI agent analysed over 740,000 on-site images, identifying more than 900 abnormal incidents. Over 2,800 early warnings were sent to prepare front-line staff ahead of the typhoon, with patrol and response data compiled afterwards and flood prevention briefings created for review. According to Zhou, the agent has “driven a fundamental shift in city safety operations, moving emergency management from reactive rescue toward proactive prediction and automated response.” Other examples included using intelligent video terminals, AI, and robotic systems to assist with routine city governance tasks. Zhou described a completely autonomous procedure for detecting and dealing with unlicensed street traders in Shenzhen. He also emphasised how AI can make public services more inclusive for all, sharing the example of a personalised chatbot agent that can help citizens navigate financial services with tailored guidance. “Technology keeps breaking city boundaries, bringing inclusive services to everyone,” Zhou concluded. Building the foundations for an AI city During the summit, Huawei welcomed customers from Angola, Saudi Arabia, Ethiopia, and China to share their experiences and advice with AI city development. Several presenters emphasised the importance of building core foundations before scaling AI. Huawei has helped more than 30 cities build this technical architecture, deploying public cloud and Ascend-powered computing as the unified base for device-edge-cloud collaboration. The city of Guangzhou in China is already two years into its journey deploying government AI applications, with 20 agents currently active and plans to reach 50 by the end of the year. Adoption is scaling rapidly. As one example, the city’s smart office agent now serves 55 government agencies and 26,000 public servants, with more than 5 billion token invocations. Nevertheless, Cheng Ou, Deputy Director-General of Guangzhou Municipal Government Services and Data Administration started her speech by explainin…