The future of the Data and AI Public Engagement Community of Practice

Between July 2024 and March 2026, we ran a community of practice for people who are or are wanting to engage and involve the public around data and AI. You can read a summary of the seventeen monthly(ish) meetings we held in a separate blogpost.

This was part of our project, ‘Giving communities a powerful say in public sector data and AI projects’, funded by Mohn Westlake, which has now come to an end. (You can find lots more resources from the wider project, such as case studies and event write-ups, here.) In this post, we reflect on the experience of running the community of practice, and what comes next.

What attendees thought

We ran a survey (well, spun up Google and Microsoft Forms) to ask what people they thought. Of 565 people on our mailing list, ten people filled out the surveys. Of those, nine had attended between one and four meetings; one had not attended any, but sometimes accessed the meeting notes. Other respondents also sometimes accessed the meeting notes (one regularly), more so than watched the videos.

Other key findings:

• We asked people how useful they found the Community of Practice for their work (out of 10), which gave an average score of 6.7
• Comments on what people found useful felt more positive: people thought the presentations were interesting and relevant (and often went beyond their own fields), valued hearing about and learning from what other people were doing, and found the links to other case studies useful. People also valued the networking element
• When asked what was less useful, several respondents thought the Q&A could be ‘a bit dead’ with fewer people on the call than expected (one answer noted Wednesday afternoon is a particularly popular meeting time and another slot might work better). Some respondents lamented that they did not have the time they would have liked to engage with the resources and the community in general. One respondent found the sessions interesting ‘but I often ponder what should I be taking away from this as a civil servant in terms of how I could help the public be more involved in issues that affect them related to data and AI’
• Eight respondents said they would come to future meetings if the community continued (the other two said ‘maybe’)
• When asked what people would keep and what they would change – several respondents said they would prefer in-person (‘People who go have really committed and are more up for discussion’ and it’s better for networking), although one said they found online events much easier to attend. One comment said they liked that the focus was ’not academic and addresses broad societal issues around AI’.

We also asked people to suggest possible new host organisations, or people they’d like to hear speak. We include them here to signpost other interesting organisations and relevant work:

• PEDRI (Public Engagement in Data Research Initiative)
• Understanding Patient Data
• Office for National Statistics
• DARE UK (Data and Analytics Research Environments UK)
• Open Data Institute
• Ada Lovelace Institute
• One HealthTech
• Civic Health Innovation Labs at University of Liverpool
• National Co-ordinating Centre for Public Engagement.

One respondent wondered if a collaborative effort – a shared secretariat across different organisations and sectors – might provide a way forward for the community of practice.

We also asked which networks we should share details with, especially inside the civil service. One answer said it would be helpful to loop in someone from the Senior Civil Service to encourage others to come along – ‘more junior civil servants can be understandably really reticent to go to public events for fear of “putting their foot in it”’, and ‘the visibility is not good about this within the CS for this sort of thing. Which is a shame.’

Our reflections

Staying on that theme, one of the things that has struck me most has been our struggle to get civil servants along to community of practice events. We have tried to plug into some relevant networks, but it’s difficult – especially from outside the civil service – to identify what these are, let alone how to work with them. As someone who helps run several other public service-facing event series (including Data Bites and TransformGov Talks), getting civil servants along to events feels like a wider challenge.

Anecdotally, there are understandable reasons for that – time pressures (especially with staff cuts), changes to working patterns post-Covid, other networks which may fulfil their needs (e.g. LGA networks). But there are also more troubling ones: the difficulties in sharing useful events with civil servants, civil servants having a sense of powerlessness to change anything having become used to firefighting in turbulent times, and damaging signals from ministers, such as the civil service speaking ban and clampdown on staff networks, which disincentivise civil servants from seeking out opportunities.

This may have affected the liveliness of the discussion after the presentations, as some survey respondents said – our first ever event was our best-attended and the discussion, on camera and in the Zoom chat, was the most flowing we had. That said, perhaps we could have structured the sessions differently to encourage others to share thoughts and insights, and taken more of a workshop approach on some occasions (as we did in December 2024).

One of my reflections a year into the project (back in March 2025) was that we were still in the relatively early stages of public servants involving the public in discussions and decisions about data and AI – perhaps we’ve not yet reached critical mass of practitioners who see themselves as such. (Relatedly: we were sometimes still grasping for the right language – public engagement? Involvement? Participation? – and the right way to talk to other people about it.)

That was reflected in some of the discussions we had in the community of practice, with lots of practical questions about how best to do things. This ranged from finding a diverse range of participants, to what methods to use, to managing people’s expectations about what the work could lead to, to how to evaluate the success of a deliberation. It felt like attendees really valued these insights, and the chance to learn from others.

My main reflection is this: there’s some really great work going on to involve the public that leads to better policies and outcomes. It’s tempting for organisations to see public engagement as an expense (and it does take time and money to do well) and a box to tick to add a veneer of trust. But the work we heard about was of real value in designing better products and policies, and supporting people to adopt new technology. Those involved were generous in giving up their time and insights to support others to run similar public exercises. There are already some brilliant case studies on how to do public involvement work well (see our other blogpost for some more detail), and a strong desire from people to learn from them and conduct their own deliberative and participatory exercises. The field may be relatively new, but it’s got off to a good start.

What next

As we said at the beginning of this post, the funding for the community of practice (and the wider project of which it was part) has come to an end. Where we had permission from speakers, we’ve published videos of presentations on individual meeting pages, on our YouTube channel and collected them in this blogpost. We’ll also be keeping the notes live for people signed up to the community of practice.

We don’t currently have any plans to run more meetings of the community of practice, and we are not handing it over to another organisation. But we hope others might pick up the mantle. You can also take a look at some of the other organisations in the field further up this post, some of whom run regular events (such as PEDRI).

All that’s left for me to say is a huge thank you – to everyone who came along and asked great questions, to the team at Connected by Data for their support, and to all of our excellent speakers. Thank you!