Oxford Insights proposes new framework to measure cities’ readiness for AI

Last Updated 42 mins ago by Kenya Engineer

Cities are increasingly deploying artificial intelligence in public services, transport, administration and other urban operations, but there is still no widely accepted framework for determining whether a city has the infrastructure, skills, governance and resilience required to deploy AI effectively.

A new study by Oxford Insights, the organisation behind the Government AI Readiness Index, is seeking to address that gap by proposing a City AI Readiness Index that would allow cities to assess and compare their preparedness for artificial intelligence.

The study, published on October 6, examined AI readiness in six cities: Abu Dhabi, Seoul, London, Montreal, Helsinki and Tallinn. Researchers found that while the cities share many of the same fundamental requirements, their approaches to becoming AI-ready differ significantly depending on their infrastructure, economic conditions, research capacity and government priorities.

The proposed index is intended to provide for cities what Oxford Insights’ Government AI Readiness Index has provided for national governments since 2017. The distinction is increasingly important as cities become the level of government at which many people directly encounter AI.

From automated public-service platforms and traffic management to planning, utilities and municipal administration, artificial intelligence is increasingly becoming part of the infrastructure through which cities interact with their residents.

Abu Dhabi provides a working example

The study takes a particularly detailed look at Abu Dhabi, where researchers examined AI readiness as it is being built and deployed rather than planned.

The UAE capital has pursued an ambitious strategy to make government services increasingly AI-driven. In January 2025, Abu Dhabi launched a three-year strategy aimed at transforming it into what the government describes as the world’s first AI-native government.

One of the most visible components is TAMM, an AI-enabled public-services platform used by approximately 4.7 million people. According to Oxford Insights, TAMM can now resolve 96 per cent of enquiries autonomously using AI.

The example illustrates why measuring AI readiness at city level requires more than counting the number of AI applications a government has deployed.

Behind an AI-powered public-service platform is a much larger technological ecosystem involving computing capacity, electricity, connectivity, data, software, cybersecurity, skilled personnel and governance systems.

Oxford Insights CEO Richard Stirling said Abu Dhabi provided an opportunity to observe the different layers of that ecosystem being assembled, from computing and energy through to AI models, investment and government adoption.

Six pillars for an AI-ready city

Oxford Insights’ proposed framework examines six broad pillars of city AI readiness.

These include the infrastructure needed to run AI systems, access to computing and energy, talent and research capabilities, financing, governance and the ability of public institutions and wider urban systems to deploy AI safely and resiliently.

The study’s emphasis on infrastructure is particularly significant. AI is often discussed primarily as a software technology, but large-scale deployment depends on physical infrastructure.

Computing facilities require electricity and cooling. AI systems depend on telecommunications networks and data infrastructure. Public-sector AI applications require secure systems for storing and processing data, while increasingly sophisticated applications can require substantial computing resources.

This means that a city’s ability to adopt AI is partly determined by conventional engineering infrastructure.

A city without reliable electricity, high-capacity connectivity, secure data infrastructure and adequate computing resources may struggle to move beyond small-scale AI experiments, regardless of how ambitious its digital strategy may be.

The same applies to human infrastructure. AI systems need engineers, data scientists, software developers, cybersecurity specialists, policy experts and technicians who can build, operate and maintain them.

Four models of AI-ready cities

The researchers identified four broad city archetypes based on how cities are approaching AI.

Platform Cities are infrastructure-led cities that focus heavily on building the technological foundations required for AI.

Cognitive Cities place greater emphasis on research, knowledge and artificial intelligence capabilities.

Delivery Cities focus on applying AI to public services and improving how governments deliver services to residents.

Strategic Cities place AI at the centre of broader government and economic planning.

Abu Dhabi and Seoul were identified as Strategic Cities under the framework.

The categorisation is not intended to suggest that one model is universally superior. Instead, it highlights that cities can arrive at AI readiness through different combinations of infrastructure, research, public-service delivery and strategic planning.

That distinction could become increasingly important as cities begin comparing their digital transformation programmes.

A city with a strong technology sector but limited public-sector deployment may have different priorities from one that already operates extensive digital government services but lacks local AI research capacity.

AI readiness is more than deploying chatbots

The proposed framework also challenges a narrow understanding of what it means for a city to be “AI-ready”.

Installing an AI chatbot on a government website does not necessarily mean that the city has developed the institutional capacity to use AI at scale.

A mature AI ecosystem requires governance mechanisms capable of determining what AI should and should not be used for, how sensitive data is protected, how automated decisions are reviewed and how residents can seek human intervention when necessary.

It also requires resilience.

As governments become increasingly dependent on digital systems, failures in computing infrastructure, telecommunications, electricity supply or software can affect essential public services.

Cybersecurity consequently becomes part of AI readiness, as does the ability to maintain services when individual systems fail.

Oxford Insights identified compute and energy requirements, talent, governance and resilience among the areas that will require continued attention as cities expand their use of AI.

A new layer of digital infrastructure

The emergence of city-level AI readiness frameworks reflects a broader change in how urban infrastructure is being understood.

For much of the modern era, city infrastructure was associated primarily with roads, railways, water systems, drainage, electricity and buildings.

Digital infrastructure subsequently became another essential layer, encompassing fibre networks, mobile connectivity, cloud services and data centres.

AI adds another layer on top of that system.

The ability of an urban government to deploy AI increasingly depends on the interaction between physical infrastructure and digital infrastructure.

Electricity networks provide power to data centres. Telecommunications networks connect users and systems. Sensors generate data. Cloud and computing infrastructure process it. AI models analyse it, while government platforms turn the resulting information into services or decisions.

For engineers and planners, this creates a new challenge: designing cities in which these systems can operate together.

Implications for African cities

The proposed index could also provide a useful framework for African cities as they seek to incorporate AI into urban management.

African cities are expanding rapidly, while governments are simultaneously trying to improve public services with limited resources. AI could potentially support areas such as traffic management, utility monitoring, revenue collection, urban planning, public-health administration and customer service.

But the readiness question comes first.

A city cannot sustainably deploy sophisticated AI systems without the underlying infrastructure and institutional capacity to support them.

This is particularly relevant where electricity reliability, broadband availability, data infrastructure and specialist skills vary considerably across urban areas.

For African cities, AI readiness may need to be approached as part of a broader infrastructure and digital-transformation strategy rather than as a standalone technology programme.

Cities could begin by identifying where their existing infrastructure is adequate, where investment is required and which public services would benefit most from AI.

The framework could also help cities avoid a situation in which highly visible AI projects are launched without addressing less visible weaknesses in data management, cybersecurity, computing capacity or skills.

From national to city-level measurement

Oxford Insights has produced its Government AI Readiness Index since 2017, providing a framework for assessing how national governments are positioned to use AI.

The organisation argues that a city-focused measure is now necessary because national averages can conceal substantial differences between individual urban centres.

A country may have a national AI strategy, for example, while individual cities differ considerably in their access to computing infrastructure, digital services, technical talent and investment.

The proposed City AI Readiness Index would provide a more granular way of examining these differences.

Oxford Insights plans to pilot the index across between 15 and 20 cities.

The expanded assessment should provide a clearer picture of whether the six-pillar framework can be applied consistently across cities with very different economic, technological and institutional conditions.

The next phase of the AI city

The development of a City AI Readiness Index comes as artificial intelligence moves from experimentation towards integration into everyday government operations.

The question for cities is increasingly not whether they will use AI, but whether they have the infrastructure and institutional systems required to use it responsibly and at scale.

Abu Dhabi demonstrates one route, combining computing, energy, investment, research and government adoption as part of a broader strategy.

Other cities may follow different paths.

Some may begin with infrastructure. Others may build around universities and research institutions. Some may focus primarily on improving public services, while others may make AI part of their wider economic-development strategy.

Oxford Insights’ proposed framework seeks to make those differences visible.

For city authorities, the value may ultimately lie less in producing a ranking than in identifying the infrastructure, skills, governance and resilience gaps that need to be addressed before AI becomes a dependable part of urban services.

As artificial intelligence becomes another layer of city infrastructure, being AI-ready will increasingly mean much more than having access to AI software. It will mean having the power, connectivity, computing capacity, skills, institutions and governance systems needed to make that technology work.