AI Strategy · October 2026
AI Readiness Assessment: A 30-Point Checklist for UAE Companies
Most AI projects that disappoint do not fail because the technology was wrong. They fail because the organisation was not ready for it: the data lived in five spreadsheets, nobody owned the process, staff were afraid of the tool, or legal discovered the project two weeks before launch. An AI readiness assessment is a structured way to find those gaps before you spend money, so that your first initiatives land on solid ground. Below is the 30-point checklist I use with companies in Dubai and across the UAE, with a simple scoring model and guidance on what to do with the result.
Readiness is the step before a roadmap. Once you know your score, the AI adoption roadmap guide shows how to turn it into a phased plan.
What an AI readiness assessment actually measures
A useful assessment looks at six areas, because AI touches all of them:
- Strategy: is there a clear link between AI and business goals?
- Data: is the information AI needs available, accurate and accessible?
- Technology: can your systems connect to AI tools securely?
- People and skills: do staff know how to use AI, and do they want to?
- Processes: are the workflows you want to improve documented and stable?
- Governance and risk: are there rules for safe, legal and ethical use?
Each area gets five yes-or-no questions. Answer honestly. "Partly" counts as no; the point is to find gaps, not to score well.
The 30-point AI readiness checklist
1. Strategy and leadership
- Leadership can name two or three business outcomes AI should improve in the next 12 months (for example, response time, cost per invoice, sales conversion).
- An executive sponsor is accountable for AI, with time and authority to make decisions.
- A budget exists for pilots, including training and change management, not only software.
- There is a shortlist of candidate use cases ranked by value and feasibility.
- Leadership agrees on what success looks like and how it will be measured.
2. Data
- The data needed for the top use cases is known and you know where it lives.
- Key data (customers, products, transactions) is mostly accurate and not duplicated across systems.
- Data can be exported or accessed through an API, not only by copying from screens.
- Someone owns each important dataset and is responsible for its quality.
- You know which data is personal or confidential and where it is allowed to be processed.
3. Technology and infrastructure
- Core systems (ERP, CRM, accounting, HR) are cloud-based or offer integration options.
- IT can provide secure access to approved AI tools with single sign-on and access control.
- There is a policy for which AI tools are approved, and unapproved tools are not the default.
- Logs and monitoring exist so you can see how tools are used.
- Internet, devices and security basics are reliable enough for daily AI use.
4. People and skills
- Most staff have used generative AI at least for basic tasks.
- Each department has at least one person interested in becoming an AI champion.
- Managers can identify repetitive tasks in their teams that AI could help with.
- There is a plan or budget for practical, role-based training.
- Staff concerns about job security have been discussed openly.
5. Processes
- The processes targeted for AI are documented step by step.
- Those processes are stable; they do not change every month.
- You can measure the current performance (time, cost, error rate) before AI.
- Exceptions and approvals are clearly defined, so it is known where humans must decide.
- Process owners are willing to redesign the workflow, not just add AI on top of it.
6. Governance and risk
- A written AI use policy exists, covering confidential data, approved tools and human review.
- Legal and compliance have reviewed obligations under the UAE Personal Data Protection Law and any sector regulator.
- There is a process to check AI outputs for accuracy before they reach customers.
- Vendor contracts address data use, storage location and model training on your data.
- Someone is responsible for reviewing incidents and updating the rules.
Scoring: where does your organisation stand?
| Total "yes" | Readiness level | What it means |
|---|---|---|
| 0–10 | Exploring | Start with awareness training, an AI policy and one low-risk pilot. Avoid large investments. |
| 11–18 | Building foundations | Run two or three focused pilots while fixing the weakest area, usually data or governance. |
| 19–25 | Ready to scale | Move proven pilots into production and set up an AI working group across departments. |
| 26–30 | Advanced | Focus on integration across systems, agents that act on workflows, and measuring portfolio ROI. |
Look at the score per area as well as the total. A company with 22 points overall but only one point in governance is not ready to put AI in front of customers, however strong its technology is.
The most common gaps I see in UAE companies
- Data in silos: sales data in the CRM, finance in accounting software, operations in Excel and WhatsApp. AI can only work with what it can reach.
- Shadow AI: employees already use public chatbots, sometimes with client data, because no approved alternative exists.
- Training as a one-off event: a single workshop raises enthusiasm for a week. Role-based practice on real tasks changes behaviour. This is the focus of corporate AI training done well.
- Pilots without baselines: nobody measured the process before AI, so nobody can prove the benefit afterwards.
- Multilingual reality ignored: teams work in English, Arabic and other languages; tools and prompts need to reflect that.
How to run the assessment in two weeks
- Day 1–2: agree scope with the sponsor and pick participants from each department.
- Day 3–6: run short interviews (30–45 minutes) using the checklist, and collect process documents and system lists.
- Day 7–8: send a short anonymous survey to staff about AI use and concerns.
- Day 9–11: score each area, map the top ten use cases on a value-versus-feasibility grid, and list blockers.
- Day 12–14: present findings and a 90-day action plan to leadership.
An outside facilitator helps because staff speak more openly and because comparisons with other organisations make the gaps clearer. If you would like support, I run readiness assessments as the first phase of my AI consultation; you can read more about what working with a consultant looks like in hiring an AI automation consultant in Dubai.
What to do after the assessment
- Fix one foundation: usually an AI policy plus approved tools, which can be done in weeks.
- Launch one quick win: a process with clear value, available data and low risk.
- Train the people involved: before and during the pilot, not after.
- Re-assess in six months: readiness changes quickly once work starts.
Frequently asked questions
What is an AI readiness assessment?
A structured review of an organisation's strategy, data, technology, people, processes and governance to identify what must be in place before AI initiatives can deliver value safely.
How long does an AI readiness assessment take?
For a mid-sized company, about two to four weeks including interviews, a staff survey, scoring and a leadership presentation.
Who should take part in the assessment?
The executive sponsor, heads of the main departments, IT, legal or compliance, and a sample of staff who would use AI tools day to day.
Do small companies need an AI readiness assessment?
Yes, in a lighter form. Even a 20-person company benefits from checking data access, agreeing an AI policy and choosing one well-defined first project.
What is the difference between a readiness assessment and an AI roadmap?
The assessment shows where you stand and what is missing. The roadmap uses that result to plan which initiatives to run, in what order, with what resources and metrics.