# Building AI Champions Across Belfast City Council Departments

URL: https://www.starien.com/blog/building-ai-champions-across-belfast-city-council-departments
Published: 2026-11-01
Author: Geoffrey Moffett
Topics: Belfast, Belfast City Council, AI training, AI champions, staff development, councils

> How a large council like Belfast could build a network of AI champions across departments in 2027, sharing safe, practical AI skills and keeping communications consistent.

## Quick answer

Building AI champions across Belfast City Council departments means training a group of interested staff from different teams to use AI safely, share practical examples and help colleagues. Champions need clear rules, regular time together and a direct line to communications, IT and information governance.

## Key takeaways

- In a large organisation, AI skills spread faster through peers than through policy documents alone.
- Champions should come from every type of department, not just the digital team.
- Shared prompt libraries and example workflows save time and keep outputs consistent.
- Champions support, but don't replace, formal governance and sign-off.

In a large council, staff in parks, leisure, planning, waste, culture, economic development and customer services are all starting to wonder how AI could help with their work. Some are already experimenting. Without support, that leads to patchy skills, inconsistent results and, occasionally, risky habits. Building AI champions across Belfast City Council departments is one practical way a big organisation could spread safe, useful AI skills quickly.

Here's how a council of Belfast's size could set up a champions network in 2027. AI tools are changing quickly, so the network itself becomes a useful way to keep everyone up to date.

## Why Champions Work In Big Organisations

Policies set the rules. Training sessions build initial skills. But day to day, people learn from the colleague at the next desk. Champions put a confident, trained person close to every team. They:

- answer quick questions;
- share examples that fit their service;
- spot risky habits early;
- feed back what's working and what isn't.

In a city council with many services and sites, that local presence is hard to replace.

## Choosing AI Champions Across Belfast City Council

Look for a mix:

- **Every type of service.** Frontline, office-based, technical and creative teams.
- **Different levels.** Not only managers. Officers often know the repetitive tasks best.
- **Curious communicators.** People who like explaining things, rather than the most technical person.
- **Volunteers with backing.** Champions need their manager's support and some protected time.

Communications, IT and information governance teams should be closely linked to the network, but the champions should mostly come from services.

## What Champions Need To Learn

### The Rules

The council's AI policy, approved tools and accounts, what data must never go in, UK GDPR basics, and the ICO's guidance and UK Government's AI playbook for the public sector as background reading.

### Practical Skills

- writing clear prompts with context, examples and constraints;
- checking outputs for accuracy, tone and bias;
- using AI for plain English and accessibility;
- summarising long documents safely;
- knowing when not to use AI.

### Communications Basics

Anything public-facing should follow the council's style and sign-off route. Champions should know how to work with the comms team, not around them.

## A Shared Prompt Library

One of the most useful things a champions network can build is a shared library of tested prompts and workflows. For example:

- "Turn this committee report into a one-page plain-English summary for residents."
- "Draft a service update for social media from this approved notice, in our style guide."
- "Suggest ten questions a resident might ask about this change."
- "Check this letter for jargon and suggest simpler wording."

Each entry should note which approved tool it's for, what data is allowed and who checked it. Review it regularly as tools change.

## Running The Network

### Monthly Meet-Ups

An hour a month, online or in person, to share wins, problems and new ideas. Keep it practical: "show us what you tried".

### A Shared Space

A simple intranet page or channel for questions, prompts and updates.

### Clear Escalation

If a champion spots a risk, such as personal data being put into an unapproved tool, they need to know who to tell, and they need to feel safe doing it.

## Examples Champions Could Share By Service

A champions network works best when examples feel close to home. In a city council, that might look like:

- **Parks and open spaces:** drafting clear notices about pitch bookings, events and seasonal opening times.
- **Leisure:** turning timetable changes into short social posts and plain-English web updates.
- **Waste and recycling:** suggesting resident questions ahead of collection changes, so the FAQ is ready before the phones ring.
- **Culture and events:** drafting event descriptions and accessible versions from a fact sheet.
- **Planning:** turning technical summaries into plain-English explanations, checked carefully by planners.
- **Customer services:** drafting replies to common enquiries from approved answer blocks.

Each example should come with the prompt, the tool, what data was allowed and how the output was checked.

## Thinking About Wellbeing And Worries

Not everyone is excited by AI. Some staff worry about their jobs, their skills or getting it wrong. Champions should be ready to listen, explain honestly what AI is and isn't being used for, and make clear that the aim is to free time for work that needs people. Leaders saying that clearly, and meaning it, makes a big difference.

## Measuring Success

Avoid counting how many people "used AI". More meaningful signs:

- time saved on specific, repeated tasks;
- fewer errors in published content;
- more plain-English and accessible outputs;
- staff confidence, measured with a short survey before and after;
- issues caught early by champions.

## A Spring 2027 Timeline

- **January to February:** recruit champions, agree time commitments, run initial training.
- **March:** champions trial three workflows each in their own teams.
- **April:** first review, update the prompt library.
- **May to June:** champions help colleagues prepare for summer, such as events, holiday service changes and visitor enquiries.
- **July:** review results and plan the next round.

## Pitfalls To Avoid

- **Making it a side project.** Without protected time, champions burn out.
- **Only picking enthusiasts.** Include thoughtful sceptics; they spot risks.
- **Bypassing governance.** Champions work within policy, not as a shortcut around it.
- **Letting the library go stale.** Assign an owner.

## How Starien Can Help

Starien is a small team based in Coleraine, working in digital marketing since 2015 with public sector teams, charities and businesses. Our [AI Coaching](/ai-coaching/) is built for exactly this: hands-on sessions that give champions practical skills, shared prompts and the confidence to support colleagues. We can also help comms teams with [Social Media Management](/social-media-marketing/) when time is tight.

We help Belfast's small businesses build AI skills too, alongside websites, social media and AEO, and [Launchpad](/launchpad/) gives owners bite-sized Playbooks and real support. [Get in touch](/contact/) to talk about training for your teams.

## Frequently asked questions

### What is an AI champion in a council?

An AI champion is a member of staff who gets extra training in using AI safely and practically, then helps colleagues in their team, shares examples and feeds back problems and ideas.

### How many AI champions does a large council need?

There's no fixed number. A practical aim is at least one champion in each major department or service area, so every team has someone nearby to ask.

### Do AI champions need technical skills?

No. The best champions are curious, good communicators who understand their service well. Training covers the AI skills; their service knowledge is what makes the examples useful.
