See it first. A sample report.
Sample report

What an assessment looks like.

This is an example of the report you receive at the end of our work. It has a summary for the board, the assessment itself, and an action plan in order of what matters most. The company and numbers are made up. The layout is real.

AI Governance Assessment

How well the company controls the AI it uses, what to fix first, and how to get there.

Example only
Sample company · example only · Prepared by Not Just 4 AI
01

Summary for the board

The company is Nearly there: 76 / 100. Rules and owners are mostly in place. What is missing is proof. Several rules are written down, but there are no records showing they are followed. Three fixes close most of the gap. With those done, the company could reach up to 88, which is Strong, within about 90 days.

02

Where you stand

Today
76/100
Nearly there
→
After the first fixes (example)
Up to 88
Strong
Not yet
0–49
Needs work
50–69
Now
Nearly there
70–84
Next
Strong
85–100

Proof you can show counts most. Written rules and owners count next. The standards that apply support the picture. How to read it →

03

What we looked at

We looked at the records below. The paperwork is mostly there. What is missing is proof that the rules are followed every day.

Policies
In place for the main areas
AI list and risk list
Kept up to date
Approval records
Partly — changes are not always recorded
Training records
In place, repeat training is late
Supplier checks
Not complete
Audit records
Available, kept differently by each team
Standards checked (supporting)
In place · 21 Partly in place · 6 Fix first · 2 · sample of 29 checks
04

What we found

  • People manage the AI tools, but no one is officially responsible when the AI makes a wrong decision.
  • The rules are written down. But the records that show they are followed are different in each team.
  • Supplier checks are the weakest area. AI tools are often bought before anyone checks the supplier.
  • Most of the standards are met on paper (21 of 29 checks). The problem is proof, not policy.
05

What to fix first

  • No one is named as responsible for the AIISO/IEC 42001
  • Suppliers of AI tools are not checkedISO/IEC 42001
  • Records are kept differently by each teamISO/IEC 42001 · NIST AI RMF
06

If someone checks you tomorrow

If a client, an investor or a partner asked today how the company controls its AI, most answers would be good enough. The first question it could not answer well is who checks the AI suppliers. Fix that first.

07

Standards and rules

On paper, the company mostly meets UAE PDPL and ISO/IEC 42001. It partly meets NIST AI RMF on measuring and monitoring. A regulator or auditor would first ask who is responsible, and look at the records. Fixing these moves the company from in place on paper to shown to work.

08

What to do

  • Name one owner for the AI, with the power to decide.
  • Check AI suppliers before signing. Also check the ones you already use.
  • Keep the same records the same way in every team, so you can show the rules are followed.
09

90-day plan

0–30 days
Name the owner; agree one way of keeping records
30–60 days
Check AI suppliers; fill the gaps in approval records
60–90 days
Review again; aim for Strong (up to 88)

Example only. Real results depend on the proof you can show and the fixes you complete.

From sample to your own

What would your report say?

Tell us in one line what you are deciding. We reply within two working days.

See also: Where you stand · Prices