Expo City Dubai says the toughest test for smart technology is governance rather than deployment

When Nadia Ferguson, the executive director of Expo City Dubai, talks about the “hardest test” for smart technology, she does not mean installing sensors, deploying cameras, or even running AI models. The real test, in her view, is what follows: can the city use the resulting data to solve real problems and make decisions that are accountable?

This question gains increasing weight as Expo City Dubai begins its experiments with artificial intelligence. Ferguson describes an organization that is still learning where technology can help and what must be in place before it can influence how the city is managed. The terminology itself is shifting: from a “smart city” that boasts technology deployment toward “urban intelligence,” where technology becomes part of the city’s operating system rather than an end in itself.

Ferguson sees three elements as a connected loop rather than separate layers: data that tells us how the city works, infrastructure that responds to that information and generates more of it, and governance that decides what can be collected, linked, and acted upon, and who remains accountable for the resulting decisions. The danger, she says, is advanced technology operating within fragmented institutional systems. The city may have massive amounts of data yet still lack intelligence.

Trust is the regulating thread for all of this. People need confidence in how their data are used, institutions need confidence in its quality, and decision-makers need to understand what the evidence shows and what it does not. The Expo City urban framework applies this approach through 131 indicators from multiple sources that assess the health of a place via environmental, social and urban outcomes.

Regarding AI, Ferguson acknowledges that they are deliberately in the early stages. The starting point is not “where we can use it” but “what we aim to improve.” So far, the most effective experiments relate to how teams work rather than how the city operates. They tested AI through teams to see where it improves quality or efficiency, where it changes thinking, and where it merely adds a layer of work disguised as productivity. They observed all three cases. This matters because the organization provides a relatively safe learning environment: a failed workflow experiment may waste time, but tolerance for error is entirely different once technology begins to affect live urban systems and people’s lives.

The clearest lesson they derived is that “AI capability” and “AI readiness” are not the same thing. A model may be technically able to perform a task, but does the organization, let alone the city, have the data, permissions, cybersecurity, governance, and human capacity to deploy it responsibly? Sometimes technology delivers the task before the institutional structure is ready. Sometimes AI saves a great deal of time. Sometimes a human completes the task faster. This friction is useful because it reveals where the true value lies.

The UAE’s ambition gives these experiments momentum. The shift toward agentic government services makes questions of infrastructure, talent, governance and institutional readiness practical rather than theoretical. Ferguson urges cities not to learn these lessons in isolation: Expo City Dubai brings cities together to compare what succeeded, what failed, and what might apply elsewhere. In her view, the next phase of urban intelligence should not be a competition over who deploys the most AI, but a question of where human and artificial intelligence combine to produce better urban outcomes.

What this means for those working in the regionThe Expo City Dubai experiment is not an academic case study. It is a live laboratory for AI governance in a Gulf urban context, and it shows that the obstacle is not model availability but the maturity of the institutions that will manage them. For those building products or policies in the region, the message is clear: investing in data readiness and human governance comes before, and outweighs, the race to deploy the newest model.