of UK small businesses using AI reported having a formal written AI policy.
10–49 employees · Fieldwork Oct 2025–Jan 2026DSIT · UK Business Data Survey 2026 ↗Your team is
already using AI.
Is your business
ready for it?_
Practical AI governance, staff training and readiness reviews for UK SMEs. Understand the risks. Give your people the confidence to move forward.
Get Good to GoAI is already
at work.
Let’s make sure
it works for you.
ChatGPT. Copilot. Claude.
They’re part of the working day.
Your people are using AI to write, analyse, research, code and make decisions. The question for your business is whether clear rules and staff preparation are keeping pace.
01 What information is being shared?
02 Can you trust the outputs?
03 Where is human oversight essential?
04 Who’s accountable when it goes wrong?
We help you answer those questions — and put the practical foundations in place to use AI confidently, responsibly and productively.
Not sure what ready looks like? Our GOOD TO GO standard sets out six practical checks for confident, responsible AI use.
of UK employees surveyed: using unapproved consumer AI tools at work.
2,003 employees · October 2025 · All employer sizesMicrosoft / Censuswide · 2025 ↗of UK employers surveyed had provided generative AI training and support in the preceding 12 months.
Autumn 2025 survey · All employer sizesCIPD · January 2026 analysis ↗Different surveys, a shared reason to act: clear rules and prepared people matter. The policy figure is specific to small businesses; the employee and training findings cover wider UK samples. A missing written policy alone does not establish unsafe practice.
“Senior staff were using AI to write Slack messages—even leaving the prompts in. Trust was slipping, and people were presenting decisions they couldn’t explain or take ownership of.”
Paraphrased from a conversation.
AI Insider
What’s really happening with AI inside UK SMEs? Anonymised stories, practical lessons and a clearer view of what good looks like.
From uncertainty.
To good to go.
Start with Discover and Assess in the Good to Go Review. The Good to Go Governance programme includes all four stages, adding practical governance, staff training and a tested handover.
Discover.
We establish what’s actually happening. Leadership interviews, a staff survey and targeted conversations uncover the tools your team uses, the data they access and the decisions AI is influencing.
Assess.
We assess significant use cases across a consistent risk framework. You get a clear view of what’s working, what needs safeguards and where to draw the line.
+ Leadership action plan
Govern.
Clear ownership. Approved tools. Data boundaries. Human oversight. We establish acceptable-use rules, escalation procedures and a practical approach to new tools — all tailored to your business.
+ AI Acceptable Use Policy
Prepare.
Practical training on using AI well: protecting confidential information, verifying outputs, understanding copyright and bias, recognising prompt injection and knowing when human judgement is essential.
+ Tested processes, follow-up and handover
Data & privacy · Security · Accuracy · Human oversight · IP & copyright · Bias & fairness · Regulatory exposure · Supplier & tool risk
Good practice.
A recognised foundation.
Aligned with key practices in the UK Government’s AI Management Essentials guidance.
See how the approach aligns
| AI Management Essentials topic | GOOD TO GO approach |
|---|---|
| AI system record (§1) | Discover: AI Use Register |
| AI policy (§2) | Govern: tailored policy and responsibilities |
| Risk assessment (§5) | Assess: prioritised risks and safeguards |
| Data management and protection (§6, §8) | Govern: data-handling rules and tool approval |
| Issue reporting (§9) | Govern and Prepare: escalation routes and staff training |
Alignment with these selected practices is our assessment. AIME is voluntary guidance, not government certification. GOOD TO GO is independent and is not government-endorsed.
Read the AI Management Essentials guidance on GOV.UK ↗Know when to go.
And when to pause.
GO
Appropriate to accelerate.
CAUTION
Permitted with safeguards.
REVIEW
Requires human or management oversight.
STOP
Unacceptable exposure.
Enough governance
to make AI safer.
Not so much that
nobody uses it.
Practical foundations. Clear boundaries. Room to move.
Set a date.
Get good to go.
Don’t let a client incident become your first AI governance review. We recommend UK SMEs put practical AI governance in place by summer 2027 — with us or another qualified provider.
Start addressing your highest-risk uses now. Set your target, assign an owner and plan the work before AI becomes harder to oversee.
Summer 2027 is GOOD TO GO’s recommended planning target, not an official or statutory deadline.
Six checks.
A clear definition of ready.
What should a business actually have in place? We’ve created a practical readiness standard for UK SMEs: six checks that define what we consider Good to Go for your team’s use of AI.
- Visible use. You know how your team uses AI.
- Clear ownership. Someone owns the decisions and risks.
- Practical boundaries. Your team knows what is allowed.
- Human control. People check the things that matter.
- Prepared people. Your team knows how to use AI well.
- Ongoing review. You keep checking and improving.
We consider you Good to Go when all six are evidenced within the agreed scope, controls work in practice and critical risks are resolved or the affected use has stopped.
Our own readiness standard and assessment, informed by recognised guidance. Each assessment records scope, evidence, remaining risks, leadership sign-off and a review date. It is not government or accredited certification, legal clearance or a guarantee of risk-free AI.

“Honestly, we weren't good to go.
The team were using different AI tools, in different ways, and important data was being uploaded in ways that were probably a little risky.
We worked with Good To Go to get a clear picture of the situation, develop some policy and governance that meant we could rein it in, and then did some follow-up work on tools that could help us run more efficiently as a business which has, in my estimation, saved us around 10% on our costs in just 3 months.
Loved how easy it was, how simple the language was, and how it made us all feel like experts instead of amateurs by the end of it!”
Ready is a good
place to start.
Understand where you stand. Put governance into practice. Find your next opportunity to grow.
Most businesses start with the Good to Go Review. It gives you the evidence and action plan to decide what happens next.
With many businesses seeking to get Good to Go, we’re now scheduling programmes for December and January.
Final pricing is confirmed after we understand your organisation and the scope of work.
Good to Go
Review
Know where you stand.
Delivered over 2–3 weeks
Understand how your team is using AI, where your business is exposed and what to do first.
- Staff survey and key stakeholder interviews
- Review of current AI tools and use cases
- Prioritised risks and gaps
- Practical guidance on immediate safeguards
- Leadership briefing and clear action plan
Clarity on your risks. Confidence in your next steps.
See how the Review works ↗Good to Go
Governance
Put confidence into practice.
Delivered over approximately 8 weeks
Build the practical foundations for responsible AI use, with clear rules, accountable leaders and a team that knows how to work safely.
- Everything in the Good to Go Review
- Tailored AI policy and data-handling guidance
- Tool approval process and clear responsibilities
- Human oversight and incident escalation procedures
- Leadership briefing and practical staff training
- Processes tested on real use cases, with follow-up and handover
Clear rules. Prepared people. Governance that works.
Explore the Governance programme ↗Good to
Grow
Turn AI potential into business performance.
Scoped around your business and opportunities
Identify where AI could save significant time, reduce costs and increase productivity—with a clear business case for where to invest.
- Discovery sessions with your key teams
- Mapping of repetitive tasks and costly bottlenecks
- Prioritised opportunities for AI and automation
- Estimated savings, implementation costs and payback
- Recommended tools, safeguards and success measures
- A phased roadmap, with optional implementation support
A practical plan for measurable business improvement.
Tell us a little about your business. We’ll confirm fit on a focused 20-minute call.
- 01 / Tell us about your businessA short form helps us understand your needs before we speak.
- 02 / A 20-minute fit callConfirm your priorities, scope, decision process and preferred start.
- 03 / Agree the work. Reserve your slot.Receive your scope and contract, then a 50% deposit invoice. Your slot is reserved when the contract is signed and the deposit received; the remaining 50% is due on commencement.
Programmes run through scheduled touchpoints, with time between them for your team’s information, feedback and approvals.
Unmanaged AI has a cost.
Do you know yours?
Repeated prompts. Unnecessary corrections. Work done twice. Small amounts of wasted time can add up—before you consider the risks to your data, decisions and client relationships.
How we calculate this · evidence and assumptions
A simple illustration.
Regular AI users × avoidable minutes per week ÷ 60 × hourly staff cost × 46 working weeks. The example uses 20 people, 30 minutes and £30 per hour: £13,800 per year in avoidable rework.
These are editable assumptions, not research-derived averages. One hypothetical incident adds 40 response hours at your hourly cost, £2,000 external support and £1,800 lost contribution (£6,000 revenue at a 30% contribution margin). At £30/hour, that is £5,000; combined with rework, £18,800 in a year with that event. The incident allowance is zero when there are no AI users.
Validate your estimate with examples from two weeks of real work. Count each correction or duplicated task once and keep essential quality checks. Exclude incident response from weekly rework to avoid counting it twice. Lost contribution is not the full value of lost revenue. Time value becomes a cash saving only if spending falls.
Why task choice and oversight matter.
A field experiment with 758 BCG consultants using GPT-4 in 2023 found suitable tasks were completed 25.1% faster on average. On one task outside its capabilities, AI users were 19 percentage points less likely to reach a correct solution. Read the research ↗
This supports assessing how AI is used; it does not establish a typical SME loss rate. IBM/Ponemon’s 2026 study of 602 breached organisations across 16 countries and regions found that 92% of those reporting an AI-related breach lacked proper AI access controls. Shadow-AI incidents involved substantial response, disruption and business costs. IBM report, findings and methodology ↗
This is research on organisations already breached, not a representative UK SME sample or proof of a particular increase in incident likelihood. It supports examining exposure; it does not set the amounts in our example. The combined figure is neither a comprehensive risk valuation nor a claim that governance prevents every loss.
Not knowing the cost doesn’t mean there isn’t one.
Get a clearer picture. Our Review maps your AI use, identifies risks and gives you a clear action plan. Governance puts the practical safeguards and training in place. Good to Grow explores better tools and ways of working.
Make your next step
a clearer one.
Explore the programmes, use the checklist and see how the work helped Kandor.
Before you get started.
What is an AI readiness review?
A structured assessment of how your team uses AI, the information involved, the risks and the safeguards already in place. The Good to Go Review gives leadership a prioritised action plan.
What does the £5,000 Review include?
From £5,000, the Review includes a staff survey, key stakeholder interviews, a review of current tools and use cases, prioritised risks and gaps, immediate safeguard guidance and a leadership briefing. Scope and price are agreed before you book.
How is Governance different?
The Review establishes where you stand. Governance, from £13,500, includes the Review and helps put tailored policy, ownership, tool approval, oversight, escalation and practical staff training into operation.
Can you help if staff use personal AI accounts?
Yes. We start by understanding actual use without blame. We identify which tools, tasks and data need attention, then recommend practical boundaries and appropriate alternatives.
How much time will our team need to contribute?
The Review runs over 2–3 weeks and Governance over approximately 8 weeks. These are elapsed delivery windows with scheduled touchpoints and time for your responses. We agree interviews, workshops, a named internal owner and response expectations in the scope; we are not embedded every day.
What does meeting the GOOD TO GO standard mean?
It means there is evidence for six practical checks within an agreed scope: visible use, clear ownership, practical boundaries, human control, prepared people and ongoing review. It is our consultancy’s readiness standard, not government certification or a guarantee against incidents.
How do we get started?
Most businesses start with the Review. Complete the short enquiry form, choose a 20-minute fit call, then receive an agreed scope, contract and invoice if the fit is right. A 50% deposit reserves the programme; the balance is due on commencement. First available programme starts are December 2026 and January 2027.


