AI Readiness Assessment: The 6 Dimensions, How to Run One, and What It Should Produce

An AI readiness assessment scores how prepared your organization is to use AI safely and get real value from it, across six dimensions: data, security and privacy, infrastructure and platforms, skills and talent, use cases and value, and governance and operating model. The quickest way to start is the free AI Readiness Score: 10 questions, about 3 minutes, and an instant score across all six dimensions, plus your red flags.

Take the free AI Readiness Score →

This guide is the hub for our AI readiness series. It covers what AI readiness means, the six dimensions, how scoring works, how to run an assessment yourself for free, what a good assessment should produce, and an honest comparison of a free score, a do-it-yourself assessment, a consultant and our paid report.

What is AI readiness?

AI readiness is whether your organization can adopt AI without creating more risk than value. It’s less a yes-or-no answer than a profile: can AI find and trust your data, can you control what it sees, do people have governed tools and the skills to use them, do you know what AI is for, and has anyone decided what’s allowed. Organizational AI readiness and business AI readiness are the same idea seen from IT and from the business side, and an honest assessment needs both voices.

Readiness is almost always uneven. An organization can have a mature cloud platform and no AI policy, or a well-trained team and badly overshared documents. A single “ready or not” answer hides exactly the information that matters, which is why a good assessment scores each dimension separately.

Why scores beat opinions

Ask five people whether the organization is ready for AI and you’ll get five answers. The enthusiast says yes, people are already using it. The security lead says no, look at the permissions. The CFO says it depends what it costs. Each is partly right, and none can settle it, because each is working from different evidence. An assessment replaces those opinions with a shared picture:

• A shared language. “We’re at level 1 on data ownership” means the same thing to everyone in the room. “Our data is kind of a mess” doesn’t.
• Specificity. Scoring forces a choice between defined answers, which exposes vague optimism and vague pessimism alike.
• Disagreements surface. When IT scores a question at 3 and a business leader scores it at 1, you’ve found something worth discussing.
• Progress you can measure. Scores can be compared quarter to quarter. Opinions just change.
• A credible answer for leadership. A board can follow six numbers and an average far more easily than a collection of views.

Scores don’t replace judgment. They separate where you stand from what to do about it, so each can be discussed on its own terms.

Why assess before you spend

The most expensive AI mistakes at mid-size organizations are rarely bad tool choices. They’re sequencing mistakes: licenses rolled out before permissions were cleaned up, a pilot funded before anyone captured a baseline, a builder hired before anyone agreed what to build. A simple question asked in advance would have caught each one. An assessment asks all of those questions at once, so the order of spending follows the order of readiness. The AI readiness checklist turns that into specific spending gates.

The six dimensions of AI readiness

A complete assessment covers six dimensions with four questions each, 24 in total. Each dimension has a detailed guide in this series.

The AI readiness framework, explained covers why these six, how the maturity ladders work and how the dimensions depend on each other.

How an AI readiness assessment is scored

Each question has five answers describing maturity levels from 0 (nothing in place) to 4 (measured, reviewed and backed by controls or automation). A dimension’s score is the points earned divided by the points possible, on a 0 to 100 scale, and the overall score is the plain average of the six dimensions. The scoring is deterministic: the same answers always give the same result. The overall score falls into one of four bands:

• Not Ready, 0 to 39: foundations are missing.
• Emerging, 40 to 59: pieces are in place, but gaps will stall pilots or create exposure.
• Ready, 60 to 79: solid foundations for targeted use cases.
• Leading, 80 to 100: mature and governed, focused on scaling value.

One rule overrides the average: if security or governance scores below 40, the rating can’t be higher than Emerging, because strong infrastructure or skills can’t offset leaking data or having no rules. How the 0 to 100 AI readiness score works goes through the math with a worked example and a scorecard you can fill in.

Turning the scores into a heat map for leadership

Six dimension scores fit on one slide. Put the dimensions down the side and, across the top, either one column for the organization or one column per business unit, then write the score and its band in each cell. Writing the band name in the cell (for example “31, Not Ready”) keeps the map readable for everyone, including colleagues with color vision differences. If you shade the cells, use a light-to-dark blue scale or blue and amber rather than red and green. The weakest row is usually where next quarter’s work goes.

What a good AI readiness assessment produces

The score is the start, not the product. Whether you run it yourself or pay for it, a complete assessment should give you:

• A scorecard across the six dimensions, with the overall score and band.
• Red flags first: answers that create exposure now, such as broadly overshared content or employees using public AI tools with work data and no controls, listed separately so the average can’t hide them.
• Ranked risks, each with its impact and a specific fix.
• Quick wins, where the next level is weeks of work rather than quarters.
• A 90-day plan, sequenced so risks close before a pilot starts, with an owner for each item.
• Budget guidance, success measures and a skills plan.
• The questions and your answers, so you can re-score next quarter the same way.

What a real AI readiness report looks like walks through a complete example page by page.

How to run an AI readiness assessment yourself (free)

You don’t need a consultant to get an honest answer. Most of what an outside assessor would collect is information your own team already has; the hard part is structuring it so it adds up. The short version:

1. Use a fixed set of questions. The 24 questions in the AI readiness checklist work as an AI readiness assessment template: each has a 0 to 4 ladder and a level 3 target.
2. Pick four to six people. The IT leader, whoever owns security, one or two business leaders, and ideally someone from finance or operations. An all-IT group tends to score the business questions too generously.
3. Gather evidence, not opinions. Sharing reports, MFA coverage, logs of AI site use, the status of your policy, the list of pilots. Give each person one or two dimensions to research.
4. Score alone, then together. Everyone answers independently, then the group discusses only the questions where answers differ. When in doubt, score lower.
5. Calculate and apply the floor rule. A spreadsheet is enough.
6. Write one page. The six scores, the band, the red flags, three quick wins and the 90-day priorities, each with an owner.

How to measure AI readiness without a consulting engagement has the full two-week schedule, the evidence to collect for each dimension, the 90-minute scoring workshop and the biases to watch for.

A preview of the checklist

Here’s one question from each dimension, with what level 3 looks like. Level 3 on every question in a dimension scores 75, inside the Ready band.

The other 18 questions, each with a guide, are in the full AI readiness checklist. If governance is your gap, the free AI acceptable use policy template is the fastest single fix.

AI readiness assessment tools and options, compared

“AI readiness assessment” covers very different products, suited to different decisions. Here’s an honest comparison, including ours.

About our report, since we sell it: the AI Readiness Report is built by David Davis. It’s written from your answers to 24 scored questions with deterministic scores, and it covers top risks rated by severity with fixes, quick wins, a 90-day plan plus a months 4 to 12 outlook, budget guidance and KPIs, and a skills-gap plan with certification paths. You get a private web report plus a PDF and one revision round, and David personally reviews each report before it’s delivered within 48 hours. Because it’s built from your answers rather than interviews, it won’t surface the organizational issues a good consultant finds in workshops; if that’s your real problem, hire the consultant. You can judge the report before buying: the full 38-page sample is public.

To go deeper on the choice, what an AI readiness assessment should cost breaks down the five kinds of assessment, what drives the price and ten questions to ask before paying, and consultant vs. assessment product vs. DIY covers when each route is right and the hybrid that often works best.

Readiness for a specific rollout or industry

The six dimensions apply everywhere, but the first risks differ by platform and by sector.

• Microsoft 365 Copilot: Copilot readiness covers Microsoft’s requirements, its built-in readiness checks and the five fixes before rollout, and fix SharePoint oversharing before you turn on Copilot is the 30-day cleanup.
• Data and cloud: data readiness for AI and cloud readiness for AI.
• Security: AI security readiness and the shadow AI policy your IT team actually needs.
• By industry: AI readiness for financial services, healthcare, manufacturing, nonprofits and professional services.

AI readiness vs. AI maturity

A maturity model describes the stages an organization moves through over years. A readiness assessment asks a narrower, more urgent question: what stands between you and your next AI step, and what should you fix first? Most mid-size IT teams should start with readiness, because it drives this quarter’s spending and rollout decisions. AI maturity model vs. AI readiness assessment explains when you need each and how to use both together.

Who it’s for

The assessment in this series is built for IT leaders at mid-size organizations who are being asked about AI and want an honest picture before committing money. Frameworks written for large enterprises assume specialist AI, data and risk teams; this one assumes a small IT team that also keeps everything else running. Bring business leaders in, too: use cases, sponsorship and value are their questions to answer.

How often to assess

Treat the first assessment as a baseline, not a verdict. Re-score every quarter during the first year, when things change quickly and the plan needs adjusting, then twice a year after that. Also re-assess before any major rollout, such as licensing an AI assistant for the whole organization, after a significant reorganization, and after any security incident involving AI. Each re-assessment is faster than the first, because the evidence sources and the framework are already familiar, and the change in scores is often the most useful thing you can show leadership.

The asset: what to gather before any assessment

Whichever route you take, these pieces of evidence make the answers faster and more honest:

1. Sharing and permission reports from your collaboration platform.
2. MFA and SSO coverage reports.
3. Web or DNS logs showing use of AI services.
4. The status of your AI acceptable use policy.
5. A list of AI tools and AI features in use.
6. Training records and current certifications for the IT team.
7. Any list of AI ideas or pilots, with owners.
8. Cloud budget and alert configuration.

And the context: your industry, size, main productivity suite and cloud, the regulations that apply, what you most want AI to do in the next year, your biggest concern, and your rough budget. Every recommendation should trace back to that context, so write it down before you start rather than reconstructing it afterward.

Where does your team actually stand?

Start with the free score: 10 questions, about 3 minutes, and an instant result across all six dimensions, with your red flags.

Get your free AI Readiness Score →

Need a document leadership will act on? The $497 AI Readiness Report is personally reviewed by David Davis and delivered within 48 hours. Flip through the full 38-page sample first.

Related guides

• The AI Readiness Checklist: 24 Questions to Answer Before Spending a Dollar
• Data Readiness for AI: The Five-Question Audit
• AI Security Readiness: Closing the Gaps AI Exposes
• Is Your Infrastructure Ready for AI? A Platform-by-Platform Check
• The AI Skills Gap on IT Teams: Measure It, Then Close It