At Skipton Building Society, organisation-wide access to Microsoft Copilot is being built around better service for its savings and mortgage customers, colleague curiosity, and clear guardrails.
In early 2026, a detailed review was carried out over several days to understand how a specific type of building address was handled within Skipton’s core banking system. In July, as part of an internal accuracy comparison, the Society’s newly developed AI-powered Q app returned the answer in just four minutes.
Q is a multi-agent retrieval tool built on Azure Foundry using GitHub Copilot by two Skipton engineers. Available as a chatbot, it helps Skipton’s business analysts, architects and developers explore the business rules in the core banking system. They can ask what a rule means, how parts of the system fit together, or where a proposed change may need to be made.
The app’s capabilities epitomise Skipton’s approach to AI. The point isn’t to hand judgement to a machine, but to give people a quicker way to investigate, learn and decide what to do next.
“We can use [AI] to create more opportunity, and we can create things that help us serve our members better“
Q can generate code and suggest solutions, but colleagues still play the critical role in deciding what problems are worth solving and how best to solve them. The technology helps people reach answers more quickly, while human judgement and an understanding of members’ needs guide the decisions that follow.
Charlie Moran, Head of Data Capability, understands the limits: “The accuracy is never going to be 100%. But my human accuracy isn’t 100% either.”
That is why human judgement remains essential. What matters, he says, is the scope it gives developers, testers and business analysts to focus on real problems rather than spending days searching through complex systems.
Purpose-led perspective
Skipton is a mutual society, owned by its members – its savings and mortgage customers – rather than by shareholders. Its purpose is to help people buy homes, save for the future and improve their financial wellbeing.
“I’m really excited to work for an organisation that has that purpose and invests in its pursuit,” says Jenny Wood, Group Chief Information Officer.
For Wood, technology choices are always made “for the benefit of our members”, she says. That helps explain why the society isn’t seeking to move everyone away from human contact: members can choose whether to interact with the 82 branches, the call centre, the chatbot or the app. Wood says the trust that comes from speaking to a person remains valued in the Society.
That lens shapes the use of AI. Rather than asking only how a tool might reduce costs, Skipton asks whether it can help colleagues solve a problem that matters to members.
“Purpose-led organisations can always ground themselves in asking what’s the right problem to solve?” Wood says.
Making room for curiosity
Wood says the society has tried to give people room to look at a problem differently. Under a cultural programme called ‘Enabling Curiosity’, colleagues are encouraged to ask whether a task should be eliminated, reduced or redesigned before deciding whether technology should automate it.
Wood says the programme is teaching the whole organisation to consider “How do you own the reinvention of your work?”
That mindset informs Skipton’s approach to Microsoft Copilot, which comes on top of a digital transformation programme involving an Azure-based Databricks data environment, infrastructure virtualisation, and a refresh of its website and apps.
Starting out, the Society gave 30 colleagues from across the business licences to explore the tool in their work. It wasn’t a formal pilot. The group compared notes, brought forward possible uses and helped the society understand where Copilot could be useful.
Wood says that work gave Skipton confidence “we had both the governance and the innovation lens in place” and then made Copilot available across its roughly 2,500-person organisation.
Adoption is strong. In July 2026, Wood says, 80% of colleagues used Copilot Enterprise.
Alongside the roll-out, the Society has continued to invest in training and capability-building, helping colleagues develop the confidence and skills needed to use AI effectively and responsibly.
Learning beyond head office
Skipton wants to give people the confidence to use AI tools well, so has encouraged adoption from staff across the country through AI training days
“This isn’t just a head office conversation,” Wood says.
“None of the work was approached from a cost-saving perspective“
The Society has run dedicated AI days for colleagues across the business, with sessions ranging from introductory guidance on Copilot to more advanced discussions around building AI-powered tools and services.
Moran says the AI deployment was never conceived as a way to do the same job with fewer people: “None of the work was approached from a cost-saving perspective. We didn’t say, ‘well, there’s 30 of you in this team, but maybe we could create the same output with 15’.”
That distinction is visible in Q. The tool can offer an answer or another option, but it doesn’t remove the need for colleagues to interpret what they find, make decisions and act upon them.
Freedom with limits
Freedom to invent comes with clear boundaries. Existing controls mean colleagues cannot put data or IT security at risk, and a change can’t be rolled out beyond 10 or 15 people without a conversation with the right team.
“You can all have approved tools, but there are rules around how you use them. And we will monitor,” Wood says.
The rules are designed to support responsible experimentation, not slow it down. Approved AI tools sit within the Society’s wider framework for data protection, governance and third-party oversight, while teams maintain visibility of emerging use cases and support colleagues in using AI responsibly.
Wood says colleagues “all want to use AI to solve a good problem for members”.
Sometimes that will mean automating a task; sometimes it’ll mean simplifying it or eliminating it altogether.
“AI is just that next thing,” she says, “We can use it to create more opportunity, and we can create things that help us serve our members better.”