Thanks. Chris will be in touch.

AI change management · UK

AI change management in the UK: how to get your team using AI every day

AI change management is the people side of adopting AI: changing habits, processes and permissions so your team uses AI safely and keeps using it. Ignite AI Solutions does it in a fixed order. Training and enablement come first, on your team's own work. Governance is the gate, so everyone knows what is allowed. Then we build Claude for Teams, our governed AI operating system, around how your team works, and run it with you in weekly sprints with ROI measured every month.

Book a SPARK Discovery Speak to Us
Why AI rollouts stall

Why do AI rollouts stall?

AI rollouts stall on people and process. The licences get bought and the tools work. Then usage fades, because the way the work gets done never changed.

“Most AI rollouts stall on people and process. It takes time to map out business processes in an organisation, and only then can it start to be automated. People need time to digitise knowledge.”

Chris Duffy, founder and Chief AI Officer, Ignite AI Solutions

The processes are not mapped yet

Claude can only take on work that has been written down. Mapping how a job really gets done, step by step, takes time. It has to happen before anything is automated.

The knowledge lives in people's heads

The templates, judgement calls and house rules that make your team good are rarely written down. People need time to get that knowledge into a form Claude can use.

Adoption is left to chance

A licence and a launch email change very little on their own. Adoption is the whole game, so it needs a plan, owners and follow-up every week.

The mindset shift is skipped

The biggest change is in how work gets done: who drafts, who checks, what waits for a person. If that shift is never discussed, people keep working the old way with a new tab open.

For the culture side in more depth, including the Human Capability Map and champion networks, read People-first AI and cultural transformation. For the common failure patterns, see why AI fails.

How we run AI change management

Three steps, each one a change in how your people work

We run AI change management in three steps: find out where your people are, train and govern them together, then build and run Claude with them week by week.

Step 1 · Discover

SPARK Discovery

SPARK runs for 4 to 6 weeks. We map how the work really gets done today, where AI pays back first and what has to be made safe.

It includes human-capability mapping across skills, leadership and culture. You find out who is ready, who needs support and where the resistance will show up. You keep the findings report.

Step 2 · Train and govern

Claude training and governance, side by side

Your team learns Claude in person, on their own work and workflows, while the rules are drawn up with them in the room. Formats run from a Team Day for up to 10 people to a Corporate Day for up to 20, plus a Leadership Strategy Day for the board.

The AI Manifesto sets the human rules. Technical governance is aligned to ISO 42001, UK GDPR, NIST AI RMF and the EU AI Act. People leave knowing what is allowed and why.

Step 3 · Build and run

Claude for Teams, run with your team

Claude for Teams, our governed AI operating system, is built around how your team already works. Each piece goes live with the people who will use it, and they are trained on it that week.

Weekly sprints keep each change small enough to absorb. Cadence meetings let leadership see progress and set the next priority. ROI is measured against the SPARK baseline and reported every month.

The enforcement layer

Safe while people are still learning

People hold back when they are unsure what is allowed. Chris puts it plainly: governance is a gate, and it removes the barriers to adoption. So we build your policy into Claude itself, as an enforcement layer that applies the rules on every task, whether or not the person asking has read them yet.

  • Drafts wait for a person before anything is sent.
  • Finance is read only.
  • Files are archived rather than deleted.
  • Anything client-facing waits for sign-off.
  • Claude works only in approved folders.

That lets your team practise on real work from day one, and it gives leaders the confidence to say yes. We follow the DSIT AI Risk Management Toolkit, published 8 September 2026. Download the toolkit.

What your team gets

Skills that stay when we go

Your team ends up able to run and improve Claude themselves. The people, the governance and the Claude setup are all yours.

Confidence on their own work

Training happens on the tasks your team does every week, so the skill carries straight back to the desk.

Rules they helped write

The AI Manifesto and guardrails are drawn up with your people in the room, so they understand them and sign up to them.

Processes written down

Work that lived in people's heads is mapped and documented, in a form Claude and new starters can both use.

A weekly habit of improvement

Sprints and cadence meetings give the team a rhythm for spotting the next job Claude can take on, and shipping it.

Numbers from your own business

ROI is measured every month against your own baseline, so decisions about what comes next rest on evidence.

Leaders who can steer it

The Leadership Strategy Day and cadence meetings keep senior people involved and visible, which is what the rest of the team watches for.

Who leads it

Who leads your AI change programme

Chris Duffy, founder of Ignite AI Solutions

Chris Duffy

Founder and Chief AI Officer

Chris's background is operations, change management and service delivery. He spent 23 years in UK Defence, 6 of them embedded with 22 SAS, and is a Certified Chief AI Officer.

Dr Joseph Ross, AI Governance Lead at Ignite AI Solutions

Dr Joseph Ross

AI Governance Lead

Dr Ross leads on AI governance. He is a Chartered IT Professional and Fellow of the BCS, holds a PhD in digital transformation and AI for the Royal Navy, and is an AI Ethics Adviser to a UK police force ethics committee.

Meet the full team

Common questions

AI change management, answered

What is AI change management?

AI change management is the people side of adopting AI: helping staff change how they work so AI tools get used safely, every day. It covers training, clear rules, process mapping, leadership support and measuring what changes. At Ignite AI Solutions it runs as training first, governance as the gate, then Claude for Teams built around how your team works.

Why do AI rollouts stall?

Most AI rollouts stall on people and process. Business processes have to be mapped before they can be automated, and people need time to get their knowledge out of their heads and into a form AI can use. When a rollout skips that work, or never talks about the shift in how work gets done, usage fades after launch.

How do you get staff to actually use AI?

Train them in person on their own work, so the first thing they see is a task they do every week done faster. Give them clear rules so they know what is allowed. Then keep going: new pieces go live in weekly sprints, people are trained on each one as it lands, and a monthly ROI report shows everyone what the change is returning.

Who can help with AI transformation and change management in the UK?

Ignite AI Solutions is a London-based AI consultancy and a member of Anthropic's Claude Partner Network as a Select Services Partner. We run AI change management alongside Claude training, AI governance and Claude for Teams, delivered by a senior team led by founder Chris Duffy. Most clients start with a SPARK Discovery.

How do you manage the people side of rolling out Claude?

Start by finding out where your people are: their skills, their confidence, and how the work really gets done today. SPARK Discovery maps that. Then train and govern together, so the team learns Claude on real work while the rules are written with them. Build Claude around their workflows one piece at a time, with regular cadence meetings so leaders stay visibly involved.

How do you keep staff safe while they are still learning AI?

We build your AI policy into Claude as an enforcement layer. Drafts wait for a person before sending, finance is read only, files are archived instead of deleted, anything client-facing waits for sign-off, and Claude works only in approved folders. People can learn on real work while the rules are applied on every task.

Does AI change management only work with Claude?

The people side applies whichever AI tool you use. We are a Claude Partner and we build on Claude, and we take no commission on licences from any vendor. If Microsoft Copilot suits your business better, we will say so.

What does our team keep when you leave?

Everything we build together. The skills, the written-down processes, the governance and the Claude setup all stay in your business. The aim is capability your team can run and improve on its own.

Next step

Start with a SPARK Discovery

In 4 to 6 weeks you will know where AI pays back first, what has to be made safe and where your people need support. You keep the findings report whatever you decide next.

Book a SPARK Discovery Speak to Us

More on transformation: AI transformation · Cultural transformation · Claude-native AI transformation · How we work · Claude vs Copilot

Chris Duffy, founder of Ignite AI Solutions

About the author

Chris Duffy Founder and Chief AI Officer, Ignite AI Solutions

Chris is the founder of Ignite AI Solutions, a Certified Chief AI Officer and one of the Top 20 AI Leaders of 2026. He is a regular expert source for Forbes on AI and cybersecurity, and a member of the UKAI Council supporting accredited pathways for upskilling. He was nominated for 3 awards at the National AI Awards 2026, including the Aiconics Award for responsible AI. A UK Special Forces veteran, he now helps UK SMEs adopt AI safely and effectively.

More about Chris · Chris in Forbes · LinkedIn · Substack

Chat on WhatsApp