The recent debate over whether Britain should have a dedicated Minister for AI raises an important question, but perhaps not the one many people are asking.
The real issue to be debated is not whether AI has a dedicated minister. It is whether governments, businesses and public services understand what it actually takes to become AI-enabled. Too often, organisations assume transformation begins by appointing a leader, creating a new department or launching a strategy. But those are only the starting points.
AI is not another function that can simply be bolted onto an existing organisation. It changes how organisations should operate, how decisions are made and how people work together. Leaders who treat it as another technology programme are likely to be disappointed. Those who redesign their organisations around it will be the ones who see the greatest benefits.
AI is an operating model, not a project
It’s not just governments either. Many organisations appear to be following a familiar pattern. They appoint a Head of AI, establish a steering committee and commission a series of pilot projects. Twelve months later they have a collection of promising demonstrations, but very little has fundamentally changed.
That is because AI does not create value simply by existing. Its value comes from being embedded in the way an organisation works every day. Processes, responsibilities and decision-making all need to evolve if AI is to move beyond isolated experiments and become part of normal operations.
I spent much of my earlier career working on the safety and production of major live events. A festival is effectively a small city built from scratch, bringing together promoters, contractors, security teams, emergency services, transport operators and hundreds of temporary staff.
You do not make a festival safer simply by appointing another safety manager while allowing every contractor, steward and supplier to continue working in isolation.
Information has to flow between teams, responsibilities have to be clear and the entire operating model has to reflect the risks on the ground. AI transformation is much the same. Appointing a leader may help, but the organisation around them still has to change.
We have seen this before with previous waves of digital transformation. The winners of the e-commerce transformation weren’t the ones that bought the most technology. They were the ones that fundamentally redesigned themselves to make the best use of it.
Leadership matters, but structure matters more
This is why appointing a Minister for AI, however welcome that may be, cannot by itself transform government.
A minister without the ability to influence how departments share data, redesign services and adopt new ways of working risks becoming a spokesperson for AI rather than an agent of transformation. The same principle applies in business. An AI leader who sits outside day-to-day operations will struggle to deliver meaningful change unless the wider organisation is prepared to adapt alongside them.
Britain’s biggest AI opportunity is adoption
Much of today’s AI conversation focuses on frontier models, investment announcements and the global race to build increasingly capable systems. Those discussions are important, but they risk overlooking where the UK’s greatest near-term opportunity actually lies.
The biggest productivity gains will come from helping organisations apply today’s AI to real operational challenges. Hospitals can use it to improve patient flow and reduce waiting times. Universities can better understand how campuses operate. Local authorities can deliver more responsive public services. Small and medium-sized businesses can automate repetitive work and make better use of limited resources.
For most organisations, the challenge is not access to AI. It is knowing how to adapt everyday workflows in ways that allow AI to genuinely improve outcomes.
We must stop treating technology as simply another part of business
There is also a broader lesson for policymakers.
Science, innovation and technology are closely connected to business, but they are not the same thing. Science creates knowledge. Innovation turns knowledge into practical capability. Business turns capability into economic growth. Strong economies understand that each stage depends on the others.
Reducing technology policy to a subset of business policy risks overlooking the ecosystem that creates future industries in the first place. Britain has long been successful at producing world-class research and entrepreneurial talent. The next challenge is creating the conditions that allow those strengths to translate into widespread adoption and lasting productivity gains.
AI will prove itself in the physical world
Perhaps the biggest misconception surrounding AI is that it is primarily a digital phenomenon.
The next phase of AI will not be defined by the number of reports it writes or images it generates. It will be defined by how effectively it helps organisations understand and improve what happens in the physical world.
That means making hospitals run more efficiently, improving transport networks, helping universities manage increasingly complex estates, reducing energy consumption in buildings and enabling safer, more responsive public spaces. These are operational challenges where better data, better decisions and better coordination can deliver tangible benefits for millions of people.
Combining digital intelligence with an understanding of the real environments in which people live, work and move will be the key to unlocking productivity, improving efficiency and ultimately driving economic success.
The question leaders should really be asking
If I could leave you with one question, it would be this: your organisation was designed for humans. What would you change if it had to work for AI as well?
That’s a harder question than it sounds. Everything about how organisations run assumes a person is doing the work: meetings where decisions live in people’s heads, processes held together by tribal knowledge, information scattered across inboxes and hallway conversations. Drop AI into that environment and it can’t operate efficiently, however capable it is. The limitation isn’t the technology. It’s the design.
We’ve been through this kind of redesign before. Accessibility taught us to adapt processes built for the majority so they work for everyone, and the changes usually made things better for everyone too. Clearer documents, simpler processes, fewer assumptions. Designing for AI works the same way. Write things down, make decisions explicit, structure information so it can be found. Organisations that do this become easier places for humans to work as well.
Whether in government, healthcare, education or business, the leaders who take on that redesign will be the ones who get real value from AI. Those who simply create another department or another title may discover that, despite all the investment, very little has actually changed.
