What is an AI readiness assessment?
It is a structured review of seven things: mission and change culture, job architecture, skills and competencies, operating model, process integrity, data and technology, and governance and risk. Each is scored, the weakest ones are named as blind spots, and the result is a sequenced plan rather than a verdict.
It exists because AI projects rarely fail on the model. They fail because the process being automated was never documented, the data had three versions of the truth, or nobody told the people whose jobs were about to change.
The seven-point readiness checklist
Answer each question honestly and count how many land on the ready side.
| Dimension | Ask yourself | Ready looks like | Not ready looks like |
|---|---|---|---|
| Mission & change culture | Can you explain, in one sentence, why AI matters to your organization? | Leadership gives the same answer, and it ties to a business outcome. | The answer is 'everyone else is doing it' or it changes by the person. |
| Job architecture | Do you know which roles change if this work gets automated? | You can name the roles, the tasks moving, and who owns the transition. | Nobody has mapped the work, so the impact surfaces after go-live. |
| Skills & competencies | Do the people doing the work have the skills to supervise the output? | There is a training plan, and time in the calendar to use it. | Training is 'we'll figure it out' or a single vendor demo. |
| Operating model | Is it clear where the human decision starts and ends? | Review, approval and escalation points are written down. | The tool is expected to 'just handle it'. |
| Process integrity | Is the process you want to automate actually working today? | The steps are documented and exceptions are the exception. | Every case is a special case and the process lives in someone's head. |
| Data & technology | Can you trust the data the system would rely on? | You know the source of record and how current it is. | The same number is different in three systems. |
| Governance & risk | Do people know what they are and aren't allowed to put into an AI tool? | There is a short written policy people have actually read. | Staff are already pasting client data into consumer chatbots. |
What your score means
0 to 2 ready
Not ready yet
Buying tooling now will cost money and credibility. Fix process and data ownership before anything gets built.
3 to 5 ready
Ready in places
There is a safe first project here. Pick the one function where the process is solid and the data is trusted, and prove value there.
6 to 7 ready
Ready to build
The groundwork holds. The risk shifts from readiness to sequencing, choosing the right first build and measuring it honestly.
How the AIIM Assessment goes further
A self-scored checklist tells you where you are uneasy. A facilitated assessment tests that against evidence, gets your leadership team to the same picture, and produces a heatmap, an explicit decision on where AI will not be used, and a sequenced 90-day plan.