How Liverpool City Region is turning AI strategy into real-world impact

An interview with Tiffany St James, Chief Artificial Intelligence Officer for Liverpool City Regional Combined Authority, UK

As part of the Government AI Campus, Apolitical is publishing a series of articles featuring leaders from the 2026 Government AI 100, a list recognising public servants around the world who are pioneering AI adoption, capacity building and regulation.

Apolitical spoke with Tiffany St James, the UK's first regional Chief AI Officer. In her role, Tiffany leads a dedicated AI taskforce, aiming to position the Liverpool City Region as a global leader in AI for Good, using artificial intelligence to improve public services, support communities and drive inclusive economic growth.

As Chief AI Officer, how do you decide which AI initiatives are worth pursuing in a city context, especially given limited resources and competing priorities?

When you're starting any programme of activity with limited resources, it's really essential to understand how you're going to focus your AI attentions, particularly within a city region. I started by looking at our strategic priorities as a combined authority, including what had been set out in the Mayoral Manifesto, our growth plan and our corporate development plan. Everything AI delivers must help those outcomes.

I then looked further afield at national doctrine, like the Modern Industrial Strategy, to consider what we're trying to achieve as a country. From there, I focused on three portfolios: AI in education, AI in health and social care, and AI in transport. I can look quite closely at those portfolios and identify what their wicked problems are. For example, we want to reduce avoidable traffic collisions that result in serious injury and death by 2040, there's a really clear parameter there. So I can then look at how AI can actually help to deliver against that particular priority.

Once you have that picture, you apply criteria: Does it deliver against strategic priorities? Could it scale beyond the region? Is there speed to market, evidence of quick wins? Is it helping the most people, including vulnerable people? Is the service team ready and is there budget?

So we start at that high level, look at strategic priorities, portfolios, wicked problems, and then put them through that criteria to come out with our shortlist.

We've also set up an AI Task Force to help steer us, which includes a group called 'Luminarees'. These are volunteers who come together for two hours a quarter and help us with our prioritisation, ideas and solutions for our AI roadmap. They come from many different backgrounds, but they can all help us change the dial on the outcomes within the region.

Can you share an example of how AI has been used to improve a public service in your city?

As a Combined Authority, we look after growth, energy and transport within the region. But it's within our purview to help, assist and support our local authorities, even though the delivery of local public services isn't directly done by the combined authority itself. So some of the programmes we've put in place are looking at where we can be adjacent to that delivery and help drive their objectives.

For example, one wicked problem is that there's a cohort of primary school children who are 10 months behind their peers in educational attainment. To address this, we've run a pilot programme where AI technology is used to create personalised learning journeys in 45 primary schools, which is around 10% of schools in the region. We want to help teachers and school governance administrators become informed consumers of AI technology and this allowed them, funded for a year, to see what difference AI and personalised learning can make. Within the last set of statistics, covering September to December 2025, we saw a 12% increase in attainment levels for children, which we're delighted with.

We also work wherever we can in what we call a triple helix: the Combined Authority, our four great universities within the city region, and local businesses, working together in a partnership. An example of that in practice is Kudata, a Korean-UK 'digital twin' programme for transport analysis. It brings together data scientists from Liverpool City Region and South Korea to look at modes of transport for more effective planning.

There's also MetaCity Liverpool, a digital twin of the entire city created by drone and photogrammetry, which gives us ingestible datasets. You can look at the cityscape and identify where solar panels would be best placed on south-facing houses, map crime hotspots using open data and monitor air quality to support better planning of movement, goods and transport routes.

How have you involved residents or frontline staff in shaping AI initiatives and what difference has it made?

It's absolutely imperative to work with frontline staff, residents and businesses. Anyone who is affected by what you're designing needs to be part of that digital design process.

We're very lucky to have had a programme called the Civic Data Cooperative, which Liverpool City Region put together in combination with the University of Liverpool and which has been running for the last three years. One of its outputs was a Data and AI Residents' Charter, developed specifically by residents, which sets out 11 principles for how they wanted their data to be managed and interacted with for future projects. That included things around responsibility, transparency and accountability.

When AI is moving so fast, it's incumbent on us to say this is how your data was processed, what it was used for, and when it will be released. We've been able to share it with other regions and with central government, who have lauded it as an example of really good practice. But this is just good digital design, ensuring that if you're putting together a programme of activity, you're including the people who will benefit from it in the design itself.

How have you used AI to support inclusive economic growth? And where do you see the greatest risks of exclusion if this isn't handled carefully?

It's really important to ensure that we're not creating an AI divide. We've had a digital divide for a very long time and we're lucky to have a dedicated team that looks at digital inclusion in the combined authority. They've always focused on helping people get online and understand the benefits of interacting digitally.

But with the rise of generative AI since 2022, keeping people safe with AI has become part of that work too. Working with the Good Things Foundation, the focus has evolved. Before, it was keeping people safe online from phishing; now it's phishing at a scale that AI can enable. But also making sure people understand that they shouldn't be putting personal information into open large language models (LLM) and that those models can hallucinate and make up responses. There's already a programme of AI literacy running for parents and residents within the combined authority.

Another portfolio I've brought to this role is capability, both within the combined authority and across our six local authorities. We must raise AI capability overall so we don't have inequalities where high-growth, highly technical businesses pull ahead and leave smaller businesses behind. Part of that is AI confidence for residents and smaller businesses. It also means making sure we have enough data from people with complex needs and from those who interact less with formal systems, so that we're not inadvertently building bias in by only using the data we already have.

The other risk is place-based inequality. Across our six local authorities, there are significant differences in economic growth and we have to make sure we're not leaving any areas behind. AI lends itself well to people who are already digitally literate and to higher-growth businesses that understand its building blocks, so we have to be really conscious that we're not compounding that existing inequality.

In your opinion, what skills or capabilities do city public servants need most when they're working with AI, and particularly in services that directly affect residents?

For me, it's about critical thinking and divergent thinking.

In the age of AI, more critical thinking than we've ever needed before. To take those results and ask: is this right? Has this left anyone behind? That responsibility sits more heavily in the public sector than in many other industries.

Divergent thinking is about ensuring you've heard from all voices, including people with opposite opinions to you, and trying to understand what has driven them to that view. Even attempting to understand an opposing point of view helps you moderate your own and leads to better results. It's the classic double diamond of design, you need to expand before you apply critical thinking.

Beyond those, everyone now needs a baseline understanding of data management and visualisation. AI is the connective tissue for data, and good, clean, interoperable data is the building block that enables faster, better decisions. That used to sit in specialist teams; now it's a responsibility across a much wider range of roles.

So over the next two to three years, how do you expect AI will change the way that city governments operate and serve the public? And what should cities start doing now to prepare?

We're moving from reactive to predictive service delivery, using AI to identify people at risk and surface interventions before they're needed. Preventing homelessness and reducing the need for temporary accommodation are two good examples.

In terms of what cities should be doing now: setting really clear priorities and building the structure to go from strategic imperative all the way through to a live case. There's a lot of focus on pilots across the sector, and we need to move from just startup to the scale-up to get real-world outcomes, not just promising experiments.

Cities also need to embed ethics now, and get it right before it becomes harder to rewind. We know that not enough diverse teams were brought into early product design historically, and there's a whole cohort of people who weren't involved, which made products less successful for everyone. The same applies to AI. We need inclusive data sets and ethical design, and we need to challenge poor practice where we see it.

LLMs are now in a position similar to social media, where the content was never the platform's responsibility. But things can change. A powerful example comes from the CEO of Beat, a UK charity focused on eating disorders in young women. After learning that OpenAI's LLM had helped teenagers hide their eating disorders from their parents, the CEO demanded change. As a result, the LLM was retrained and it now responds by asking whether the user might have a disorder and directing them to support charities. It's an illustrative story that we all have the power, when we see practice that isn't ethical, to challenge and have changes made. That is a huge responsibility for cities and the public sector.

Finally, collaboration is utterly key. Rather than isolated solutions, we need regional ecosystems, sharing what we're learning and building on each other's practice. Our mayor went to see Michael Bloomberg in New York as far back as 2017, and we continue to learn from city states further abroad as well as each other within the UK, sharing what we're trialling and blueprint-setting here in Liverpool City Region so others can use it too.

What excites you most about governments using AI, and what concerns you?

What excites me most is the ability to solve real-world problems, not just within our region, but to blueprint those solutions for other city regions in the UK and further abroad. We're at this wave of ingenuity and responsibility, and that's a really exciting prospect.

We need to be better at working with other organisations, partners and regions because we're more powerful together. We're also lucky to have the Hartree Centre in Daresbury with a supercomputer up the road, offering a slice of compute at a reasonable cost to even smaller businesses in the region. So looking at our sovereign AI capability is, of course, really exciting.

The concern is the speed of change, and whether our systems and processes are agile enough to keep up. This isn't just about AI, it's about change management across the way we procure, the way we hire and the way we collaborate, to make sure we can adopt and lean into new capabilities faster than we have before.