How AI-Ready Is Your Business? 10 Questions to Find Out

Before I start any AI strategy engagement, I ask the leadership team to rate their own AI readiness out of ten.

The average answer: seven or eight.

Then I run the actual assessment.

That gap between perceived readiness and actual readiness is the most consistent finding across every business I work with. It’s rarely down to complacency; most leadership teams are neither wrong nor careless about it. It’s that “AI readiness” isn’t the same thing as “people using AI tools occasionally.” It’s a far more specific set of capabilities, and most businesses have never sat down and mapped them.

Here are the ten areas we assess. Work through them honestly and you’ll have a clearer picture of where your business actually stands, and where the gaps are.

1. Do you have an AI strategy, or just AI activity?

These are not the same thing. AI activity is “some of our team use ChatGPT.” An AI strategy is a documented set of decisions about which tools you’re using, for which purposes, who owns AI adoption across the business, what success looks like, and how you’ll measure it.

Most businesses have activity. Fewer have strategy. Without strategy, AI adoption stays uncoordinated, inconsistent, and impossible to scale.

Green: Written AI strategy, reviewed by leadership in the last 6 months. Amber: Informal direction exists but isn’t documented or consistently communicated. Red: No strategy. Tools adopted ad hoc, no clear owner across the business.

2. Do your people know what AI tools they have access to?

Sounds like a low bar. It isn’t.

A Finance Director at a professional services firm told me recently, without hesitation, that they had four AI tools. Their IT team came back with twelve. The gap wasn’t rogue spending. It was AI quietly embedded in existing SaaS products through updates the business hadn’t noticed: CRM systems, marketing platforms, project management tools. All of them had added AI functionality in the last 12 months. Almost none of it was being used on purpose.

Green: Complete, current inventory of AI tools across the business, including those embedded in SaaS products. Amber: Core tools are known; embedded AI tools in SaaS products not fully mapped. Red: No inventory. Leadership doesn’t know what’s running.

3. Are your data permissions fit for purpose?

This is the question that most AI rollout reviews fail. Your data permissions were probably last reviewed when your team was a different size, in a different structure, doing different work entirely.

Copilot and similar tools surface information at a speed no manual search can match. If your permission structure is out of date, AI doesn’t stop at what you meant it to find. It finds everything it can reach, including files that shouldn’t be accessible, sitting in structures that made sense two org charts ago.

Before you expand AI access to more of your team, your data governance needs to match its current state, not the state it was in when it was last configured.

Green: Data permissions reviewed in the last 12 months, with AI access specifically in scope. Amber: Known gaps but no remediation plan in place. Red: Permissions haven’t been reviewed since the tools were set up.

4. What percentage of your team uses AI at least weekly?

Regular, intentional use is the foundation of everything else. You can have strategy, tools, and permissions all in place, and if actual adoption stays low, none of it compounds.

We use weekly usage as the threshold because occasional use rarely produces meaningful productivity change. The savings come from habit, from AI becoming the first thing someone reaches for when a task arises rather than the last.

Green: 60% or more of the team uses AI tools at least once a week for work tasks. Amber: Pockets of adoption, inconsistent across teams. Red: Adoption confined to a small group of enthusiasts; most staff aren’t engaging.

5. Can your people articulate what AI is useful for?

There’s a question I ask at the start of every training session: “What would you use AI for in your job this week?”

The answers tell you everything.

High-readiness teams give specific answers immediately. “I’d use it to draft the proposal that’s due Thursday.” “I’d use it to analyse the spreadsheet that came in from the client.” “I’d use it to prep my agenda for the board meeting.”

Low-readiness teams give vague answers. “Writing?” “Research maybe?” “I’m not really sure what it can do.”

Use-case fluency, knowing specifically what AI can do for your particular job, is the difference between a tool people reach for and a tool people quietly ignore.

Green: Most staff can name three specific use cases for AI in their own role without prompting. Amber: Some staff have working use cases; others don’t. Red: Staff don’t have clear use cases. AI feels abstract rather than practical.

6. Do you have a clear AI governance framework?

Who decides which AI tools can be used, and for what? What’s the process for requesting a new AI integration? What happens if someone uses AI in a way that creates a data risk?

Governance doesn’t mean restriction. It means clarity, and clarity means your people can use AI confidently, knowing what’s in bounds, rather than avoiding it because they’re not sure where the line sits.

Green: Written AI governance policy, communicated to staff, reviewed in the last 12 months. Amber: Informal norms exist but aren’t documented. Red: No governance framework. People are making individual calls about use.

7. How are you measuring AI’s impact?

If you’re not measuring it, you’re not managing it, and you’re not making the business case to the board for continuing to invest in it.

Measurement doesn’t need to be complicated. Hours saved per person per week, tracked through simple self-reporting, gives you a directional number meaningful enough to act on. Licence utilisation data from your AI provider gives you visibility on adoption. Post-training assessments tell you whether the skills are actually landing.

Green: Defined metrics for AI adoption and impact, tracked at least quarterly. Amber: Some tracking, not systematic. Red: No measurement. AI investment assessed on gut feel at best.

8. Is your leadership team visibly using AI?

AI adoption tends to follow the waterline of leadership behaviour. If the Managing Director isn’t using AI, isn’t talking about it positively, and isn’t asking questions about it in team meetings, adoption stays sluggish regardless of what the official policy says.

The businesses with the fastest adoption rates are consistently the ones where the MD or CEO is a visible, enthusiastic user. Not necessarily an expert, but an engaged learner, happy to be seen figuring it out in front of everyone else.

Green: Leadership team actively uses AI, talks about it in team contexts, and models the behaviour they want to see. Amber: Some leaders engaged, others indifferent. Red: Leadership isn’t using AI and isn’t visibly invested in the business doing so either.

9. How dependent are you on a single AI vendor?

The events of June 2026, when the US government required Anthropic to suspend access to Fable for non-US citizens overnight and without warning, raised a question most UK businesses hadn’t thought to ask.

What happens if your primary AI tool becomes unavailable?

It doesn’t need to be a government action. It could be a pricing change, a service outage, a security incident, or a policy shift. If your business is entirely dependent on a single AI platform, that’s a concentration risk your IT governance should already be accounting for.

Green: Multi-tool strategy with at least two primary AI vendors; continuity plan documented. Amber: Awareness of the risk; no formal mitigation. Red: Single-vendor dependency with no awareness of, or plan for, disruption scenarios.

10. When did you last run an AI skills assessment for your team?

AI capabilities move fast enough that training from 18 months ago isn’t fully current any more. The tools have evolved. The use cases have expanded. The skills your team needs to be effective in 2026 are genuinely different from the skills they needed in 2024.

A skills assessment isn’t only about identifying who needs upskilling. It’s about understanding where your team actually is, so you can target development at the areas that will produce the highest return.

Green: AI skills assessed in the last 12 months; development planned against the findings. Amber: Assessment done but not recently; no follow-through on development. Red: No skills assessment has been done.

Your score

Count your greens, ambers, and reds.

7 to 10 green: You’re in a strong position. Focus on measurement and continuous improvement. 4 to 6 green: Solid foundations with meaningful gaps. Prioritise governance, measurement, and adoption. 0 to 3 green: You’re earlier in the journey than you might have thought. That’s not a criticism. It’s the most useful thing you can know right now, because it tells you exactly where to start.

The point of this assessment was never to make businesses feel behind. It’s to create the clarity that strategy actually requires. You can’t plan from a position of comfortable ambiguity.

If you’d like to run a proper AI readiness assessment with your leadership team, get in touch. We use a structured ten-question tool that produces a scored output in under three minutes, and a gap analysis that gives you a prioritised action plan.

Matt Neal is the founder of Artificia1. We help UK SMEs develop genuine AI capability, from assessment and strategy through to training and implementation.