AI readiness and maturity assessment for your organisation
Before picking AI projects, assess four areas with a few yes, in-progress or no questions each: strategy, data and AI infrastructure, talent, culture and skills, and responsible AI. Score yes as 2, in progress as 1 and no as 0, have several leaders fill it separately, and fix the weakest area first, because a project that depends on that area will stall there.
Why assess first
Companies start AI projects without the data access or the people to finish them, and then blame the technology when the projects stall. Accenture's research on AI maturity found only 12% of the organisations it studied qualified as "AI Achievers", so most companies are earlier than they think.
Questions
| Area | Questions |
|---|---|
| Strategy | Do senior leaders actively champion AI initiatives? Is there a documented AI strategy tied to business goals? Is AI work planned, or reactive to trends? |
| Data and AI infrastructure | Can teams access the data they need, with permissions handled? Are there pipelines and a data team? Can a team experiment and deploy without months of setup? |
| Talent, culture and skills | Are there people who can build and run AI in production? Do employees have access to AI training? Is experimentation encouraged and shared? |
| Responsible AI | Is there an ethics and governance process? Are risks reviewed before launch? Is there human oversight for high-risk uses? |
Scoring
Each question scores 2 for yes, 1 for in progress and 0 for no. Total each area and compare across areas rather than adding everything into one number, because one weak area can block everything else.
| Area score (out of 6) | Reading |
|---|---|
| 5 to 6 | Ready to scale use cases in this area |
| 3 to 4 | Pilots possible, fix gaps in parallel |
| 0 to 2 | Fix this before starting projects that depend on it |
Run it with several leaders
Ask three or more leaders to fill it in separately, then compare. The biggest disagreement is usually more useful than any score, because it shows where leaders see the company differently.
What weak areas usually mean
- Weak strategy: projects chase trends and get stopped when sponsors change.
- Weak data: AI projects fall over at the first real dataset, so hire or assign data engineers first.
- Weak talent: pilots get built by vendors and nobody can run them afterwards.
- Weak responsible AI: projects stall in legal review late, or launch with risks nobody checked.
From the floor
I built this assessment as a spreadsheet for leaders in the AI foundations and business alignment session, and when a participant asked whether every organisation should run it, I said yes, because it shows where you are: in a company of a thousand employees, some people might be working with generative AI on their own, with no company work behind it and no access to AI training, and answering these questions shows you're at an early stage without the strategy, data, talent or responsible AI practice to implement a project.