Urban technology has advanced rapidly over the past decade, with cities investing heavily in sensors, connectivity, and digital platforms to improve services. These initiatives have delivered meaningful progress, but they have also revealed structural limits: As cities become more complex, they are increasingly focused on transforming existing data into coordinated, real-time action.
Smart cities taught us how to connect urban systems and make city operations more visible through sensors and integrated data. Cities are now looking for ways to strengthen these foundations with more intelligent tools that enhance decision-making and service delivery. Rather than adding more devices, the focus is shifting towards building the capability to act at speed, and towards open, cross-domain platforms that support co-creation across industry, government, and academia. In this context, AI becomes a natural extension of the smart city framework, enriching and supporting existing strategies rather than replacing them.
What we refer to as “AI City” describes this AI-enabled operational capability within smart city development, sometimes framed as Smart City 4.0 or AI City 1.0, to reflect the role of AI as an enabling layer. An AI-enhanced smart city uses real-time decision-making and orchestration, with models continuously predicting demand, optimising resources, and coordinating services.
Smart cities are largely about sensing and integration, helping them see what is happening. AI builds on these capabilities by adding deeper contextual understanding, predictive insight, and coordinated response across services. Together, these capabilities help cities transform data into consistent, real-time action.
From insight to coordinated action for reliable, scalable AI
A key challenge for AI-enabled cities is ensuring that data can be translated into coordinated action at scale. As urban systems become more complex and interdependent, cities need more than analytical insight; they require decision-making processes that can anticipate conditions, evaluate possible responses under constraints, and support execution across multiple services.
This shift involves moving beyond static reporting or descriptive analytics towards more dynamic forms of decision support. Cities must be able to interpret signals from across infrastructure, mobility, utilities, and public safety systems, and understand how these signals interact under changing conditions such as demand fluctuations, infrastructure stress, or environmental events.
Equally important is the ability to evaluate interventions within real-world constraints. Urban decisions are rarely made in isolation; they must account for safety requirements, regulatory frameworks, resource limitations, and competing service priorities. This makes coordination across systems and departments as important as the underlying analytical capability itself.
Finally, decision-making does not end at analysis or recommendation. Cities need mechanisms to ensure that decisions are carried through operational workflows and that outcomes are continuously monitored. This creates a feedback loop in which cities can learn from real-world results and progressively refine both models and operational policies.
When these elements are aligned — data interpretation, decision evaluation, and execution — cities are better able to manage complexity in real time while maintaining transparency and accountability.
What this looks like in practice
In practice, the most visible applications of AI in cities are emerging in areas such as mobility, infrastructure management, public safety, and environmental resilience. These domains are often where the gap between data visibility and real-world outcomes becomes most apparent.
In mobility systems, cities are exploring ways to improve traffic flow, parking efficiency, and transport coordination through more adaptive and responsive approaches. This reflects a broader shift from systems that only monitor conditions towards systems that can also anticipate demand and support real-time decision-making. In infrastructure, predictive methods are being used to identify early signals of system stress and reduce the likelihood of service disruption.
In public safety and environmental management, AI-supported monitoring is being used to improve situational awareness and response coordination. Similar approaches are emerging in energy management, water systems, and urban environmental monitoring, where real-time conditions and resource constraints require more coordinated decision-making.
Across regions, including cities in Asia and Europe, these developments point towards a consistent direction: Urban systems are gradually shifting from isolated digital tools towards more integrated operating models, where sensing, decision-making, and execution are more closely connected.
Across these developments, the focus is increasingly operational rather than technological. The emphasis is on how cities interpret signals, how decisions are made under constraints, how actions are executed across systems, and how outcomes are evaluated to improve future performance. This requires stronger coordination between systems, clearer governance structures, and continuous feedback between operations and decision-making processes.
