There's a difference between "our business has access to AI tools" and "our business is ready to use AI effectively." Lots of companies fall into the first category. Fewer are genuinely in the second. The gap isn't technical. It's organisational.
UK businesses increasingly have access to AI — but readiness remains the constraint. 54% of UK SMEs report using AI on the British Chambers of Commerce’s 2026 measure, while the ONS’s official figure for all UK businesses was 29% in June 2026. Part of that gap is methodology; part of it is firms “using AI” on paper but not in the operating model, with usage that spans everything from well-governed pilots to uncontrolled shadow adoption. The bar is also moving from chat to agents that execute multi-step workflows, and MIT’s Project NANDA found that 95% of custom enterprise generative AI pilots fail to reach production — so readiness now also means narrowing scope hard and proving one workflow end to end before scaling. Being “ready for AI” doesn’t mean you need expensive infrastructure. It means clarity: knowing what you’re trying to solve, how you’ll measure success, and who owns the outcome.
October 2026 update — a baseline you may already have. One item on the checklist below asks whether you can measure if AI is actually improving things. If your business uses Copilot Cowork, part of that measurement is now supplied. Microsoft’s Copilot consumption dashboard reports “Cowork assisted hours and value” across eight task types, and Microsoft has published the methodology behind it. It is still rolling out, so check Message Center before assuming your tenant has it. Treat the figure as a conservative proxy to baseline against, not a number to hand your finance team. At the readiness stage its use is simple: record it before a pilot starts, alongside your own timing of the process you intend to change, and you have a “before” to compare against. Data can also now be exported from the Copilot and Agent 365 dashboards, so that baseline can sit in your own records.
Notice that "have invested in expensive AI infrastructure" isn't on the list. Readiness is about understanding and process, not technology shopping.
Genuinely ready businesses can name it: "We process 150 invoices per month, each one takes 20 minutes." If you can't describe it specifically, you're not ready yet.
AI needs data. It doesn't need perfect data. But it needs to be accessible. Can you get the data to the AI tool when you need it?
Someone has to define the rules: "If the invoice is over £10k and from a new supplier, flag it." These rules come from people who actually do the work.
If you're comfortable with "let's test this on 10% of the volume first, then expand," you're actually ready.
The businesses that succeed are ruthlessly specific. Start narrow and prove the model before expanding.
AI can't codify something that isn't actually codified. Sort out the process first.
If the underlying issue is that your team doesn't have a clear process, AI will just automate the inconsistency.
If you can check at least six of these, you're genuinely ready to pilot an AI solution.
One item has hardened since this checklist was first written: governance. UKSI 2026/425 has been in force since 12 May 2026 and requires the ICO to produce a statutory code on AI and automated decision-making, expected in 2027, and the Data (Use and Access) Act 2025 raises accountability expectations. Before any pilot, confirm a named AI-risk owner and DPIA coverage. It is a gate, not a nice-to-have.
A note on quick wins vs. capability building: You can get fast value from a narrow AI project. But real business transformation comes from building AI-readiness across your leadership and team. The quick wins are how you build the internal case for that larger capability.
First: Choose a specific process and understand it. That clarity is 80% of the work.
Second: Get the data accessible.
Third: Define the policy and rules.
Fourth: Find the pilots.
Book a 60-minute conversation. We'll work through the checklist, identify gaps, and create a roadmap to AI readiness — no pressure to buy anything.
Schedule a Conversation