The short answer
An AI readiness assessment checks whether your business can actually get value from AI yet. It measures five things: skills, real use-cases, data handling, governance and leadership backing. It does not measure how much software you have bought. Most businesses discover their gap is skills rather than tools, which is the cheapest and fastest gap to close. You can run a useful version yourself in about ten minutes using the ten questions below.
What an AI readiness assessment actually is
An AI readiness assessment is a structured way of answering an awkward question: if we spent money on AI next month, would any of it stick?
It is not a technology audit. Nobody needs to inspect your servers. The honest version looks at your people, your work and your rules, because that is where AI projects succeed or quietly die. A business with two ChatGPT licences and a team that uses them daily on real work is far more ready than a business that has signed a large enterprise deal nobody has opened.
It is also not an AI maturity model in the consultancy sense, where you are placed on a five-stage ladder and sold the next rung. A readiness assessment should end with three or four specific things to do, in order, starting this month.
The most common finding is not that a business is unready. It is that it is ready in some areas and completely unprepared in one, and that one area is holding everything else up.
The five things a readiness assessment should measure
Readiness is not a single number. These five dimensions each fail differently, so it is worth scoring them separately.
| Dimension | What it really asks | What failure looks like |
|---|---|---|
| People and skills | Can your team prompt well, check output and spot a confident wrong answer? | Tools are available but usage is occasional and nobody trusts the results |
| Use-cases | Have you named the actual tasks, by team, where AI would save real hours? | Enthusiasm about AI in general, with no specific job it is doing |
| Data and safety | Do people know what must never be pasted into a public tool? | Client data in a free consumer account, and nobody knows it happened |
| Governance | Is there one page of rules, and someone who owns the decisions? | Everyone improvising, or a blanket ban that pushes usage underground |
| Leadership | Does the top of the business use it and ask about it? | Approved in a meeting, then never mentioned again |
A ten-question AI readiness check
Score each question 0 for no, 1 for partly and 2 for yes. Answer as things actually are, not as you would describe them to a client.
- Is at least one person using AI every day on real work, not as a novelty?
- Could you name three specific tasks, in three different teams, where AI would save hours each week?
- If two colleagues were given the same task, would they produce comparably good results?
- Does everyone know which information must never go into a public AI tool?
- Do you have a written AI policy, even a single page?
- Is there one named person who decides which tools are approved?
- Has anyone had practical, hands-on AI training rather than watching a webinar?
- Do people check AI output against a source before it reaches a client?
- Do you know roughly how many hours AI is saving the business at the moment?
- Has a director or owner used an AI tool on their own work in the past month?
What your score means
| Score | Where you are | The right first move |
|---|---|---|
| 0 to 6 | Early. Usage is accidental and unmanaged. | One tool, one team, three real tasks, and a half day of hands-on training |
| 7 to 12 | Patchy. Keen individuals, no shared standard. | A one-page policy plus team training so good habits spread beyond the enthusiasts |
| 13 to 17 | Working. Real value, uneven coverage. | Close the weakest dimension and start measuring hours saved |
| 18 to 20 | Genuinely ready. AI is part of how work gets done. | Deeper skills: agents, automation and tool-specific advanced training |
For context, most UK businesses that run this honestly land in the 7 to 12 band. If that is you, you are in the middle of the pack rather than at the back of it. Our free AI capability assessment does the same job interactively in two minutes and gives you tailored next steps at the end.
Four mistakes that make readiness assessments useless
- Counting licences instead of habits. Seats bought is a procurement number, not a readiness number. Ask how many people used a tool yesterday.
- Asking only the enthusiasts. The person who volunteers to answer is the person already using AI. Ask the sceptic in accounts too, because their answer is the real baseline.
- Scoring ambition rather than practice. "We plan to" is a zero. Readiness is about what already happens on a normal Tuesday.
- Stopping at the score. A number with no prioritised actions attached changes nothing. The output that matters is the list of what to fix first.
What to do with the result
A readiness assessment earns its keep in the fortnight after you run it. A sensible sequence:
- Fix the lowest-scoring dimension first, not the most interesting one. An unaddressed data-safety gap outranks an exciting automation idea.
- Write the one-page policy if you do not have one. It takes an afternoon and removes most of the risk people worry about. Our AI usage policy template is a starting point you can adapt.
- Train on your own real work. Generic training produces generic enthusiasm; practical sessions on your actual tasks produce output people use the next morning.
- Pick one number to watch, usually hours saved per team per week, and check it a month later. Measuring the ROI of AI training is easier than most people expect once you have a baseline.
- Re-run the assessment after a quarter. Readiness moves fast once training starts, and the second score is usually the encouraging one.
Readiness is not a gate you have to pass before you are allowed to start. It is a map of what to do next. Most businesses that feel unready are actually one policy page and one good training session away from being genuinely capable, which is a much smaller distance than it feels from the inside.
AI readiness assessment - frequently asked questions
What is an AI readiness assessment?
An AI readiness assessment is a structured check of whether your business can actually get value from AI yet. It looks at skills, real use-cases, data handling, governance and leadership rather than which tools you have bought, and it ends with a prioritised list of what to fix first.
How long does an AI readiness assessment take?
A useful self-assessment takes about ten minutes for a single view and a couple of hours if you gather answers from a few team leads. Anything sold as a multi-month readiness audit is usually a consultancy project wearing a questionnaire as a disguise.
What should an AI readiness assessment measure?
Five things: whether your people have practical skills, whether you have identified real tasks worth automating or assisting, whether your data handling is safe, whether someone owns the decisions, and whether leadership is genuinely behind it. Skills and use-cases carry the most weight.
Do I need an AI readiness assessment before training my team?
Not necessarily, but it makes the training far better value. Knowing which tasks matter and where the skills gaps sit means a session can be built around your real work instead of generic examples, so people leave with something they use the next morning.
What is a good AI readiness score?
Most businesses score in the lower half, which is normal rather than alarming. The score matters far less than the pattern: a low skills score with strong leadership support is a very different problem from good skills with no governance, and each needs a different first move.
Which Oxford AI School training fits this need?
Choose practical training around your team, tools and real work. Sessions are available in Oxfordshire, London, across the UK, in the United States and internationally through live online delivery.
Get your readiness score in two minutes
Our free AI capability assessment benchmarks your team and gives you tailored next steps. Or get in touch and we will talk it through with you honestly.
Take The Free Assessment
Harry Lang runs The Oxford AI School, teaching teams and businesses the basics of AI in a fun, easy and practical way. A marketing leader turned AI trainer, he can be reached at Learn@TheOxfordAISchool.com.