AI Readiness Assessment: A 20-Question Checklist

An AI readiness assessment checks whether your company can use AI well before you spend on it. The checklist below has 20 yes or no questions across five areas: plan, data, rules, people and projects. Every “no” is a task, and the area with the most is where to start.
The whole checklist is on this page, with no download or sign-up. For help acting on the answers, see the consulting page or my fractional Chief AI Officer service.
What is an AI readiness assessment?
It is a structured look at whether your company has what it needs to get value from AI. That usually means a clear plan, usable data, rules for safe use, trained people, and a way to move projects from test to daily work. If the plan is the weak spot, start with an AI strategy and a 90-day roadmap.
Readiness is not the same as maturity. Readiness asks whether you can start or scale now. A maturity model describes the stages a company grows through over time. The assessment gives you the starting line. Maturity is the map of what comes after.
Why check before you spend?
Because most companies are less ready than they think. Cisco’s 2025 AI Readiness Index surveyed more than 8,000 senior IT and business leaders at companies with over 500 employees. It found that only about 13% qualified as “Pacesetters,” the group that beats its peers on every measure of AI value. Cisco says that share has not moved in three years. Cisco’s press release (October 14, 2025) shows where the gap sits:
| Practice | Pacesetters | All companies surveyed |
|---|---|---|
| Has a defined AI roadmap | 99% | 58% |
| Has a change-management plan | 91% | 35% |
| Has a mature, repeatable process for scaling AI use cases | 62% | 13% |
| Tracks the impact of AI investments | 95% | Not stated. Cisco says Pacesetters do it at about three times the rate of others |
Source: Cisco AI Readiness Index 2025. The figures are practices reported by the leaders surveyed. They do not prove that these practices cause better results.
The pattern is useful anyway. The leaders are not winning on tools. They are ahead on plans, process and measurement, and those are all things the checklist below asks about. To check whether AI is paying off once you start, see AI ROI.
The 20-question AI readiness checklist
Answer each question yes or no. If you cannot answer it, or two leaders would answer differently, count it as a no. This is a working tool, not a validated scoring model, so it gives you no total score and no pass mark. The count of “no” answers in each area tells you where to look first.
Area 1: Is there a plan?
| # | Question | Yes / No |
|---|---|---|
| 1 | Can your leaders name the top three AI priorities without checking notes? | |
| 2 | Is there a written AI roadmap, with owners and dates? | |
| 3 | Is there a budget set aside for AI, apart from individual tool subscriptions? | |
| 4 | Does one named person own AI across the company? |
Area 2: Is the data usable?
| # | Question | Yes / No |
|---|---|---|
| 5 | Do you know where your customer, finance and operations data lives? | |
| 6 | Is that data stored in a few systems, not scattered across spreadsheets and inboxes? | |
| 7 | Does someone own data quality, including who can see what? | |
| 8 | Could you safely give an AI tool access to one real data set this month? |
Area 3: Are there rules?
| # | Question | Yes / No |
|---|---|---|
| 9 | Is there a written AI policy that staff have read? | |
| 10 | Is there an approved-tools list, and a way to ask for a new tool? | |
| 11 | Do staff know which kinds of data must never go into an AI tool? | |
| 12 | Could you answer a customer, investor or board question about AI risk today? |
Area 4: Are people ready?
| # | Question | Yes / No |
|---|---|---|
| 13 | Has every person who uses AI at work had some training? | |
| 14 | Do you know which tools staff are already using, approved or not? | |
| 15 | Are managers able to say which of their team’s tasks AI could take over? | |
| 16 | Is there a plan to tell staff how AI will change their jobs? |
Area 5: Do projects reach daily use?
| # | Question | Yes / No |
|---|---|---|
| 17 | Can you name one AI project that reached daily use? | |
| 18 | Does every AI project have a named business owner, not only a technical one? | |
| 19 | Does every project have a success measure set before it starts? | |
| 20 | Is there a rule for when to stop a pilot that is not working? |
What do the benchmarks say about each area?
Each area below has a sourced number so you can see how common a “no” is. Survey figures describe the groups surveyed, not your company.
Plan. Cisco found that 58% of all companies have a defined AI roadmap. Microsoft and LinkedIn’s 2024 Work Trend Index found that 79% of leaders agree their company must adopt AI to stay competitive, yet 60% worry their leadership lacks a plan and vision for it. See the Work Trend Index.
Data. Gartner’s survey of 1,203 data management leaders, run in July 2024, found that 63% of organizations either do not have or are unsure whether they have the right data management practices for AI. Gartner also predicted that through 2026, organizations will abandon 60% of AI projects that are not supported by AI-ready data. Read the Gartner release (February 26, 2025). A forecast is a forecast, so treat the 60% as a warning, not a measurement.
Rules. For how to build the policy itself, see the guide to AI governance for a mid-size company. ISACA’s 2026 survey of digital trust professionals found that only 38% of organizations have a formal, comprehensive AI policy. Another 30% have a limited one and 25% have none. See ISACA’s release.
People. The Work Trend Index found that 78% of AI users bring their own AI tools to work, and only 39% had received AI training from their company. ISACA found that 33% of organizations train all employees on AI. If your “no” answers cluster here, my AI workshops are live, hands-on training for teams.
Projects. IBM’s 2025 study of 2,000 CEOs found that only 25% of AI initiatives had delivered the expected return, and only 16% had scaled across the whole enterprise. See the IBM study.
How do you read your answers?
Count the “no” answers in each of the five areas. There is no magic total. Use this instead:
| What you see | What it usually means | First move |
|---|---|---|
| Most “no” answers are in Plan | Nobody senior owns AI | Name one owner and write a one-page roadmap with three priorities |
| Most are in Data | Projects will stall on access and quality | Pick one data set, assign an owner, clean it, check the systems and cloud setup can connect to it, then run a pilot |
| Most are in Rules | Staff are using AI with no guardrails | Write a one-page policy and an approved-tools list |
| Most are in People | Adoption will lag the tools | Train the people who already use AI, then each team on its own tasks |
| Most are in Projects | Pilots start but do not finish | Pick the two pilots worth finishing and stop the rest |
| Roughly even across all five | The gap is leadership, not one function | Start with Plan, since the other four need an owner |
If “no” answers pile up in Plan, or are spread evenly, the fix is rarely a tool. It is a person with authority to decide. That is the job a fractional Chief AI Officer takes on, and the guide to what a Chief AI Officer does describes it.
Should you do it yourself or hire someone?
A self-check is enough when you have a small team, a clear owner and honest answers. Have two or three leaders answer separately, then compare. Differences between answers are often more useful than the answers themselves.
An outside review earns its cost in three cases:
- Your own leaders keep disagreeing about the answers.
- A customer, investor or regulator is asking for evidence.
- You have the answers but nobody with the time and authority to act on them.
A good outside assessment should end with a ranked list of gaps, an owner and a date for each one, and a short roadmap. A slide deck full of scores and nothing to do on Monday is a warning sign. If the gaps are about direction, AI strategy consulting shows what you should hold at the end of an engagement.
When I start with a company, these are the questions I ask: How are you using AI today? What tools are you using? Does anyone share skills with the company? Is there a second brain where everyone shares information with each other’s AI agents? What are your biggest pain points, and what is the most time-consuming manual work you want to automate?
If the gaps point to a missing AI leader, compare the options in when to hire a fractional Chief AI Officer, fractional vs full-time and the market cost data.
How often should you repeat it?
Every three to six months, and after any big change such as a new tool rollout, an acquisition or a new rule from a customer or regulator. This interval is a practical suggestion, not a research finding. Keep the old answers so you can see which “no” answers turned into “yes.”
What are the next steps?
- Run the 20 questions this week with two or three leaders, answering separately.
- Circle the area with the most “no” answers.
- Name one owner and one date for each of the first three gaps.
- Run the checklist again at your next quarterly review.
If you want an AI leader to own the gaps, my fractional Chief AI Officer service covers the plan, the policy and the training. If you want a second opinion on your answers first, talk to me about consulting.


