Looking beyond automation and cost savings

For public service leaders facing significant financial pressures, the appeal of AI-driven efficiencies is understandable. But David encourages organisations to think more broadly about what the technology could make possible.

"I don't think people are going to be saying, oh, it saved us a lot of money."

Instead, he sees opportunities to fundamentally rethink how people interact with public services. AI could make services more accessible through conversational interfaces, speech-to-text technology and personalised information. Rather than requiring people to navigate complex websites and processes, organisations could explore ways to make information and support easier to access and understand.

The challenge, David suggests, is bringing together an organisation's purpose, the capabilities of AI and the creativity to imagine services working differently.

Learning from AI implementation in government

David's experience includes helping develop one of the UK's first citizen-facing applications of large language models for the Foreign, Commonwealth & Development Office (FCDO).

The project began as an opportunity to experiment with generative AI, but with an important condition: any experiment should have the potential to become a real service.This meant addressing the practical questions that accompany AI deployment, particularly in environments where inaccurate information can have serious consequences. For example, providing incorrect advice about visa requirements or medication restrictions could have significant implications for travellers.

The discussion explores how organisations can approach these risks proportionately, combining rigorous testing, safeguards and access to existing services where AI cannot provide an appropriate response. Antonio also raises the importance of understanding existing performance before introducing new technology.

"Does technology help me make this thing better, sufficiently better than how it was before to justify any costs and risks that come with it?"

Without a clear baseline, organisations may struggle to demonstrate whether AI has genuinely improved a service.

Moving from experimentation to meaningful change

As AI tools become increasingly accessible, many organisations are already experimenting with them. But individual productivity improvements do not necessarily translate into organisation-wide benefits.

David argues that public service organisations should begin with a clear understanding of their purpose, rather than the technology itself. This requires a connection between strategic priorities, the experiments organisations choose to pursue and the people responsible for implementing them.

It also means investing in staff capability. While generative AI tools can appear straightforward to use, developing the judgement and skills to apply them effectively is more complex. Antonio reflects on The PSC's work with organisations to develop AI portfolios that bring together experimentation, governance and strategic direction.

The discussion highlights the need to balance leadership and coordination with the freedom for teams to explore opportunities – avoiding both excessive central control and disconnected experimentation.

What will AI mean for the future of work?

Looking ahead, David considers how AI could change not only the tasks people perform but the nature of their roles.

He points to software development, where AI is increasingly taking on coding tasks while developers focus more on reviewing outputs, making architectural decisions and overseeing systems. Similar changes could emerge across other professions, with people spending more time directing and evaluating work completed by AI.

But this raises an important challenge: how will people develop the experience and judgement required for more senior roles if AI performs the tasks through which they would traditionally learn? David is particularly concerned about the implications for education and early-career development. Organisations will need to consider how to use AI while preserving opportunities for people to develop their own knowledge and skills.

Designing the future we want

Throughout the conversation, David and Antonio return to the importance of human agency. AI capabilities are developing rapidly, and their eventual impact remains uncertain. But uncertainty does not mean organisations should simply wait to see what happens. David draws on an idea that has influenced his thinking:

"The future is best not thought of as a forecasting exercise, like weather, but as a design challenge."

For public services, this means making deliberate choices about where AI can add value, how risks should be managed and what role people should play in shaping its development.

The opportunity is not simply to introduce new technology into existing services, but to consider how those services could work better for the people who rely on them.

Watch on YouTube or listen to the full episode of The PSC in Conversation to hear Antonio and David discuss the opportunities, risks and practical challenges of AI adoption – and what it could mean for the future of public services.

The PSC exists to make public services brilliant. If you are interested in finding out more about how your organisation could adopt AI safely and responsibly, get in touch with Antonio Weiss.