Cities Know What to Do. Why Can’t They Do It
There is no shortage of strategy in cities today. Artificial intelligence strategies, climate transition plans, and digital roadmaps are everywhere. The language is clear, the ambition is there, and in many cases, the direction is not in question. And yet, progress often feels slow, fragmented, and uneven. This is not a failure of vision. It is a question of capability.
A growing body of research, including work by the UCL Bartlett School of Planning through its Public Sector Capabilities Index, suggests that the way we evaluate public sector performance needs to shift. Instead of focusing on what institutions plan to do, it asks a more direct question: do public institutions actually have the capability to deliver? This reframing matters because it highlights a gap that is often overlooked—the distance between intention and execution.
Cities today are not lacking solutions. Technologies are evolving rapidly, private sector innovation is accelerating, and examples of good practice are widely shared across Europe and beyond. In many cases, the ecosystem surrounding cities is moving faster than the cities themselves. Companies are ready to deploy, tools are available, and knowledge is accessible. Yet implementation remains difficult. The reasons are not always visible from the outside, but they are consistent: procurement processes that are not designed for emerging technologies, limited internal technical capacity, fragmented governance structures, and decision-making environments that are shaped by risk rather than action. These are not strategic challenges. They are operational ones.
Artificial intelligence is bringing this gap into sharper focus. Moving from AI strategy to real deployment requires cities to manage and govern data, work across institutional silos, engage with external partners, and make decisions under uncertainty. None of this is solved by having a well-written strategy document. It requires internal capability—teams that understand the technology, processes that can adapt, and leadership that is willing to move despite incomplete information. Without this, strategies remain aspirational.
Some cities are beginning to address this challenge more directly. Rather than producing new strategies, they are focusing on how to align internal processes, build capacity within their teams, and test new approaches in practice. These efforts are often incremental and rarely visible at scale, but they point to an important shift—from planning to doing. They also highlight how difficult that transition actually is.
This is where most urban innovation conversations fall short. The focus tends to remain on solutions, technologies, and best practices. While these are important, they rarely address the harder question: what actually enables a city to move from knowing to doing? Without that, even the most advanced ideas remain theoretical, disconnected from the realities of implementation.
If cities are to deliver on their ambitions—whether in artificial intelligence, climate transition, or public services—the conversation needs to evolve. It needs to move from asking what cities should do, to understanding what enables them to act. This shift is not about lowering ambition. It is about grounding ambition in the reality of how cities function.
Cities are entering a phase where expectations are higher than ever. Citizens expect better services, technologies are advancing at speed, and policy frameworks are becoming more demanding. In this context, the ability to act becomes the defining factor. Not the number of strategies produced, but the capacity to turn them into outcomes.
As part of the broader AI & the City exploration, this raises a fundamental question: how can cities build the capability to move from strategy to implementation? Because the future of urban governance will be defined by what they are able to deliver.
