A Citizens’ Assembly in the sunny hills of the Bannau Brycheiniog National Park

by Ellen Coughlan and Elinor Laws

🌱🌳🌲🌳🌱

At Green Man festival this summer, we turned the Pandemonium stage into a citizens’ assembly. The idea of a citizens’ assembly has been gaining ground in recent years as a way of deliberating tough decisions on big, complex issues. Instead of leaving it to politicians or experts, a group of ordinary people come together to listen, learn, and deliberate. Assemblies in the UK and Ireland have shaped decisions on everything from climate change to abortion law, showing that when we come together with knowledge, trust, and an open mind, we begin conversations that produce illuminating, thoughtful, and honest insights.

We borrowed a little of that spirit at Green Man by asking festival-goers to weigh in on two issues of health equity and data science: should we unlock more data for personalised medicine and prevention, and should the NHS invest in AI right now? Our aim wasn’t to solve these issues once and for all, but to get people thinking and talking and to bring together perspectives that don’t often share the same space.

The first motion asked whether the NHS should be able to use data from every aspect of our life, for instance shopping, housing, transport, genetics, on an opt-out basis, to improve prevention and tailor care.

Those in favour argued this could be transformative, spotting risks earlier and filling gaps in population health data. The opposition countered that we should invest first in people, services, and trust. Why build vast, risky datasets when immediate inequalities like access to housing or local care remain unaddressed? The Noes won by a landslide. Cybersecurity fears loomed large, as did scepticism and distrust in the private sector. People also raised questions about data quality and privacy. Some audience members, though, were cautiously enthusiastic: perhaps a pilot with 100,000 volunteers could test the waters, and maybe there could be real benefit if these data are used to represent communities that don’t often show up in routine NHS data.

The second motion asked whether the NHS should invest in AI now, or pause activity until the trade-offs are clearer and technologies more equitable.

Supporters argued that the technology is already here and can ease the administrative burden on clinicians while improving decision-making. They saw AI as an inevitable and necessary step forward. Opponents warned that AI is no shortcut to fairness. Without the basics in place, including equity of access, fit-for-purpose infrastructure, and robust oversight, AI could do more harm than good. Some worried it’s being forced into the system too quickly, without space for reflection or safeguards.

This time, the Ayes carried the day. The audience recognised that AI isn’t going away, and that it could help the NHS do more with limited resources. But concerns lingered about how fairly it could be rolled out, how ready hospitals really are, and whether clinicians risk losing skills by relying too heavily on algorithms.

Reflections

What fascinated us most was how the format shaped the conversation. By setting up the debates as binary choices, we asked people to take sides, even though a reasonable position could (and perhaps should) sit somewhere in the middle. That tension produced some of the richest discussions. Our goal was curiosity over consensus, and the citizens of Green Man delivered.

It was inspiring to see such deep engagement from people of all ages and professions. Some were experts; for others, it was their first time wrestling with these ideas. We were especially grateful to hear the perspectives of young people. Their voices too often left out of debates on data and AI, even though the consequences are shaping their futures. 

Hosting a citizens’ assembly was also a powerful way of demonstrating the wide-ranging and varying views amongst people. It is easy to assume that our own views are widely shared but it is events like citizens assemblies that remind us that this often isn’t the case. We tend to live in echo-chambers, surrounding ourselves with people who share similar world-views, and online algorithms feed us the opinions we want to hear. It is important that we as researchers and policy makers are open to hearing the other side of the story, and consider how our choices of ‘who is in the room’ can influence our overall perception of public opinion.

In the end, the festival stage reminded us what a citizens’ assembly is really about: ordinary people grappling with extraordinary questions, and reminding us that the future of healthcare isn’t just about what technology can do, but about what kind of system we want to build together. 

Since emerging from the mountains and returning to our daily lives, we have been reflecting on those conversations and connections and thinking about ways to integrate this powerful participatory research method into our work at DSxHE.

One step we are taking is to set up a Patient and Public Involvement working group to design and guide our work. We are looking for people with experience in PPIE work and strategy, including people with lived experience!

If this is something that might interest you, please get in touch with us at info@datascienceforhealthequity.com