Artificial intelligence is already changing the ways cities work. From LA to Verona, AI is easing traffic congestion. Meanwhile in Liverpool it is being used to track shopper footfall, optimising city flows.
But these examples are only scratching the surface of what the technology can do in the concrete jungle.
As AI becomes embedded in the fabric of urban living, it has the potential to fundamentally reshape how we navigate and engage with urban environments. But there’s an important conversation to be had not just about where we use the technology in cities, but how it will be implemented, who will benefit. And ultimately, how do you keep citizens onside?
Only through positive experimentation with the technology, implementing universal design principles and a strong commitment to fostering public trust will we begin to understand how AI will optimise the user experience (UX) of navigating and being in a city for residents and visitors. AI will be a lot like the internet: it will just be there – a ubiquitous part of our experience of everyday life, and not something you’d need to set up a profile for.
And if AI is going to be as all-encompassing as the internet, we have to work out how to apply it in a way that works for everyone.
As simple as getting from A to B?
One of the easiest ways to explain what AI is and how it works is with wayfinding. Most people have already used a mapping app that changes the recommended route according to real-time conditions.
Smarter offerings – suggested destinations and more efficient routes that show people not just where to go but what to look out for on the way – are low-hanging fruit for the future of wayfinding.
Google Maps – by far the most commonly used navigation app – is already providing this tailored information to its users thanks to the inclusion of Gemini, the Google AI platform. In time, this process will become even more sophisticated: navigation apps will be able to anticipate when you might want to travel – and provide precise suggestions that account, for example, for the fact there’s a marathon taking place in the city that day.
However, having lots of apps on your phone acting as different data points is potentially frustrating from a UX perspective: digital fatigue is a well-documented phenomenon. A unified way of delivering a personalised experience is needed, and super-apps – the likes of Grab, for example – present a neat solution to this problem.
But the technology will have an even greater impact behind the scenes.
By integrating with existing infrastructure, such as transport APIs, AI can analyse patterns and pre-empt failures. A subway carriage prone to overheating might be flagged before it causes a breakdown. Traffic signals can adapt in real-time to prevent congestion without the commuter ever noticing.
So, as is the case with Verona’s smart traffic solution, AI’s impact will be felt by people without them even knowing it: minimising service disruptions and extending the lifespan of publicly funded assets.
But while this invisible optimization is impressive, it can leave people uneasy. 87% of Brits support laws requiring AI systems to be proven safe before release, and only 9% trust tech CEOs to regulate AI in the public interest.
If AI is everywhere, but nowhere to be seen, how do we hold it to account?
Winning resident buy-in
In order to unlock the benefits that come with AI, residents will have to buy into its potential. That means having a strong incentive to surrender their data, including their whereabouts. In short – a clear and obvious value exchange.
Consumers have proved in other circumstances that they’re happy to share personal data if what they get in return seems like a good deal. Think of store loyalty points, or a health-tracking app.
If AI-driven systems and, indeed, cities, truly improve people’s lives, then citizens and visitors will be happy to cooperate by sharing their data.
A Deloitte survey from late last year indicates that consumers are happy with their doctors using generative AI to convey information about new treatments (71%), interpret diagnostic results (65%), and even diagnose conditions and illnesses (53%). But crucially, these statistics hinge on patients being able to trust that their data is being handled correctly.
People would be more willing to share their data if the value exchange is clear, the protections robust and the purpose well-communicated. For example, citizens would be willing to accept AI-powered systems that can cause a car to brake in an emergency when it detects a pedestrian or a dog has run into the road. Or a system that times the delivery of their kerbside pickup from a store to coincide with their arrival at the meeting point.
But the line between accepting AI for its convenience, and feeling overwhelmed by its constant presence is hard to draw. The solution is to build a system that residents can trust.
Trust won’t come from convenience alone. People need assurances that their data is being protected, and an explanation of why it’s needed in the first place in order to make new AI systems more effective – something that will only happen through clear, transparent communication. Only then will citizens be able to trust the benefits that come from integrating AI into urban living.
The danger of catering to the average
Absolute trust (if that is even possible) in AI won’t come unless people know its application has everybody’s interests at heart.
It sounds logical that we should aim to implement AI in a way that works for most people. Catering to the average. But the solution that will grow people’s trust in AI will make sure that it works for everybody.
Inclusive systems are, quite simply, better systems. A mapping tool that can identify, for instance, a tricky set of stairs, might be designed with commuters with disabilities in mind, but would also benefit elderly citizens, parents with toddlers, and tourists who just have weary legs.
If it is going to be all-encompassing, but not seen as invasive, then the integration of AI cannot be approached in any other way than with universal design principles at its heart. It’s the only way to solve the trust issue.
And, especially while AI’s role in daily life is still in its infancy, human involvement is more important than ever. Human designers bring empathy, creativity, and a nuanced understanding of diverse user needs, which automated tools can’t replicate.
From a design perspective, skepticism towards AI in general can actually be useful. Feedback from concerned parties is incredibly useful – it helps designers address concerns from the outset of the design process, and address factors like bias, or misuse, for example.
Human understanding is still needed, not just to make AI solutions feel trustworthy, but to make sure users experience the city at its best. There’s more to city living than getting from A to B as fast as possible.
Imagine being able to optimise not just for speed (which may be a commuter’s priority) but for the first-time visitor’s best experience. A tourist on a trip of a lifetime might appreciate that within a short walk of their next stop is a historic site. Imagine being able to optimise not for speed, but for enjoyment.
The best possible journey for the individual won’t necessarily be the best ‘average’ journey.
Preserving character
Cities are made special by their unique quirks, rhythms and inconsistencies. They have souls. A day in London feels very different to a day in Paris, New York, or Barcelona, and smart AI systems have to be designed to preserve the character of cities.
In the same way brands are using AI to tailor messages to “an audience of one,” cities can use AI to create tailored experiences. Imagine a system that nudges you toward a hidden art installation or flags a local flea market as you stroll past. AI can be about joy, not just logistics.
Sure, they can already help people optimise a journey for speed, but for richness of experience, tailored to the individual and the place, that takes some pretty intelligent AI. With the right value exchange for consumers and service providers, AI is going to revolutionise the urban experience for all.
