EP22: Reviewing the AI Opportunities Action Plan: What are the Opportunities Today?

As the UK moves from AI ambition towards implementation, the relationship between technology, government and the public is becoming increasingly important. In this panel discussion, Baroness Thangam Debbonaire brought together perspectives from healthcare, public policy, law and technology to explore how the UK can accelerate AI adoption while maintaining public trust, accountability and a meaningful role for people.

Moving Beyond AI as a Separate Topic

For Indra Joshi, a doctor and health technology leader with experience across the NHS, government and startups, the most important change over the next two years would be to stop treating AI as a standalone topic.

In healthcare, AI should instead be viewed as an underlying enabler of broader transformation. Rather than creating separate strategies around AI itself, organisations should focus on the problems they are trying to solve and consider where AI can contribute.

This distinction is important beyond healthcare. The goal should ultimately be for AI to become embedded into everyday services rather than remaining something that is constantly discussed as a technology in its own right.

AI Is About to Become More Visible in Everyday Life

Dr Elizabeth Seger, Director of Digital Policy at Demos, argued that the next six months could see AI becoming increasingly tangible for the public. People are likely to encounter AI through transportation, public services, politics and collective decision-making, meaning that questions around responsible deployment will become much harder to ignore.

This could create both an opportunity and a risk. As AI becomes more visible, public expectations around trustworthy regulation are likely to increase. Concerns about automated decision-making are already emerging in conversations around areas such as welfare and lending.

The challenge is ensuring that public trust does not become an afterthought once AI is already embedded in people's lives.

Closing the Gap Between Innovation and Regulation

From a legal perspective, Alex Kirkhope, a technology partner and head of an AI advisory practice, highlighted the continuing uncertainty around the UK's future AI regulatory framework.

While recent legislation has addressed aspects of data and technology regulation, there remain unanswered questions about what a more comprehensive approach to AI will look like.

For businesses, greater clarity over the UK's direction of travel would provide an important foundation for investment and innovation. This does not necessarily mean that a single AI Bill is the solution to every issue, but businesses need to understand how the regulatory landscape is likely to develop.

Using AI to Improve Public Services

Ed de Minckwitz, European Director of Public Policy at ServiceNow and formerly Head of Public Sector AI Adoption in the Cabinet Office, focused on the potential for AI to address some of the UK's most pressing public-sector challenges.

Backlogs in areas such as immigration, courts, GP appointments and hospital appointments represent both political and citizen-facing problems. Applying successful private-sector approaches to these challenges could provide some of the clearest demonstrations of AI's practical value.

The ambition should not simply be to introduce AI into government, but to use it to improve the experience of citizens interacting with public services.

Bringing Politicians Into the AI Conversation

A recurring challenge is that policymakers cannot make effective decisions about technology they do not understand.

Dr Seger described a six-month AI parliamentary scheme designed to help parliamentarians understand the fundamentals of AI, including how systems work, where data comes from and what their inputs and outputs are. The purpose was not to promote particular policy positions, but to create enough technological literacy for politicians to ask better questions.

This kind of education could become increasingly important as AI moves into more areas of policy and legislation. Politicians cannot be experts in every subject they regulate, but they need sufficient understanding to assess the strategic, economic and social implications of new technologies.

Using AI to Strengthen Digital Democracy

AI could also help politicians engage more effectively with the public.

Participatory democratic processes traditionally require significant amounts of time and money, particularly when large numbers of people are involved. Dr Seger highlighted the potential for AI-supported digital participation to allow larger groups of people to contribute to complex policy discussions more efficiently.

The key, however, is closing the loop. People need to see how their contributions have influenced the final outcome if these processes are to build rather than undermine public confidence.

AI could therefore become a tool not just for delivering services, but for strengthening the relationship between citizens and government.

From Policy Documents to Practical Demonstrations

For Ed, one of the biggest gaps between industry and government is that businesses often explain what their technology does without demonstrating how it can solve a government's specific problems.

AI creates an opportunity to change this dynamic because organisations can develop and test relatively small pilots much more quickly than traditional technology transformations.

Giving ministers and policymakers hands-on experience with functioning AI systems could help bridge the gap between abstract policy discussions and practical understanding. The shift from scan, pilot, scale could allow government to experiment with solutions, demonstrate their value and build confidence before committing to larger-scale deployment.

The Public Needs a Reason to Care About AI

The panel also explored a deceptively difficult question: how do you explain the benefits of AI to someone who is not interested in the technology itself?

For Dr Seger, the answer is partly that AI is already affecting people's lives. It influences what people see online, how information is presented to them and increasingly how decisions are made.

That means the public conversation cannot focus solely on hypothetical future risks. People need to understand both the opportunities AI offers and the ways it is already shaping their everyday experiences.

A Better Connected Public Sector

Indra offered a different perspective on what AI could mean for citizens, arguing that some of the most exciting improvements are not necessarily AI problems at all.

Better-connected public services could allow information to move between systems more effectively, making it easier for citizens to access the services they need. AI can help enable these changes, but fundamental issues around data standards, interoperability and information governance still need to be addressed.

The lesson is that government should not automatically reach for AI when the underlying problem is simply that systems do not communicate effectively.

Democratising AI Through Personal Agents

For Indra, one of the most exciting developments is the growing ability for individuals to build their own AI agents.

She described creating an agent to manage different calendars, appointments and commitments — effectively bringing together information from different parts of everyday life. As these tools become easier to create, AI could move from something people consume through large technology platforms to something they actively configure around their own needs.

This represents a broader shift towards the democratisation of AI, where individuals can use increasingly accessible tools to solve practical problems in their own lives.

Liability and the Complexity of AI Systems

The panel returned repeatedly to the question of liability.

Modern AI systems are often built from complex technology stacks involving multiple organisations, models and providers. Determining responsibility when something goes wrong can therefore be considerably more complicated than with traditional software. A clearer distribution of liability could provide businesses with greater certainty while also helping consumers understand where responsibility lies.

For the UK, this could become an opportunity to use its reputation as a leading legal and financial centre to help establish internationally credible approaches to trustworthy AI.

Building an AI Assurance Industry

The UK's existing reputation around AI safety could also provide an economic opportunity.

Dr Seger argued that the UK should consider building a major AI assurance industry, with companies specialising in model testing, evaluation and assessment. Rather than seeing regulation and safety solely as constraints on innovation, the UK could turn its expertise in these areas into an exportable service.

This could allow Britain to differentiate itself not through completely separate AI rules, but through becoming a trusted place to test, evaluate and deploy AI responsibly.

Sovereignty Through Collaboration

The question of AI sovereignty also featured prominently in the panel's closing discussion.

The UK is unlikely to compete with the largest technology powers by attempting to replicate their entire infrastructure independently. Instead, collaboration could provide a route towards greater technological independence.

Open public infrastructure, shared data and international collaborative projects could help countries outside the largest AI ecosystems maintain greater control over their technological future.

For the UK, this would complement its existing strengths in research, regulation and AI assurance.

Keeping Humans at the Centre

The final discussion returned to perhaps the most fundamental question: where does the human fit into an increasingly automated world?

Baroness Thangam Debbonaire emphasised the importance of understanding where human beings remain essential as AI changes the nature of work and creativity. This is particularly important for creative industries, where questions around copyright, ownership and the ability of creators to continue producing work are becoming increasingly pressing.

The objective should not simply be to determine which jobs AI can replace, but to think more carefully about the kinds of work, creativity and human contribution that society wants to preserve and expand.

Building an AI-Ready UK

The panel's discussion ultimately pointed towards a broader definition of AI readiness.

The UK needs regulatory clarity, but also better public-sector systems, stronger digital infrastructure, greater AI literacy among policymakers, effective liability frameworks and meaningful public participation.

At the same time, the country has an opportunity to build on its existing strengths in law, AI safety, research and innovation.

As Baroness Thangam Debbonaire concluded, the challenge is not simply to make AI work faster. It is to decide where humans should remain at the centre of the system, how government can keep pace with technological change, and how the UK can build a future with AI that benefits the people it serves.