Insights
How to choose an AI course for executives in Egypt and the GCC
Choosing an AI course for executives in Egypt or the GCC comes down to nine questions, and price is not the first of them. This guide, written by Ahmed Amrousy — Executive Education Instructor at the American University in Cairo and founder of Menova — sets out what to ask a provider about delivery language, hands-on ratio, who actually stands in the room, and how confidential data will be handled.
Last updated:
Most organisations in Egypt and the Gulf are now on their second or third conversation about AI training. The first one usually produced an impressive day and very little change. This is a guide to making the next one produce something.
There are nine questions worth asking. Price is not the first of them.
1. Who will actually stand in the room?
The single most common disappointment in corporate training in this region is buying a name and receiving an associate.
Ask directly: who delivers the session? What have they run outside a training room? An instructor who has held budget responsibility argues differently from one who has only ever taught — because they have had to defend a decision that cost money.
2. What language will it genuinely be delivered in?
“We can deliver in Arabic” covers three very different realities: a trainer who thinks in Arabic; a trainer working through an interpreter; and English slides read aloud with occasional Arabic asides.
For most teams in Egypt and the Gulf, the format that works is verbal delivery in Arabic with English slides and materials. Participants ask real questions in the language they think in, while the written material stays in the language of their systems.
Ask which of the three you are being sold.
3. What proportion is hands-on — and what will people actually produce?
Ask for the agenda and count. If more than a third is presentation, the team will enjoy it and change nothing.
The sharper version of the question: what will each participant have produced by the end of the day? A good answer is specific — three rebuilt workflows, a redesigned assessment, a one-page AI position. A vague answer means slides.
4. Will exercises run on our own data?
Generic exercises are safe, easy to prepare, and forgettable. Working on the organisation’s real documents is the difference between a team that understands the idea and a team that has already done the work once.
This has a prerequisite, which is question five.
5. How will the data boundary be handled?
If a provider proposes running exercises on your real material without first establishing what may and may not be entered into a public AI tool, that is disqualifying.
In regulated sectors — energy, financial services, government-related entities — the boundary session should come first, before any exercise touches real material. Done properly it does not limit the training. It is what makes using real material possible at all.
6. Does it teach tools, or judgement?
Tool training has a shelf life of about a quarter. Interfaces change, features move, and the provider’s carefully prepared screenshots go stale.
What does not go stale is the ability to brief a model properly, to ground it in a real source, and to tell when the output is wrong. Ask what the participant will still be able to do in eighteen months.
7. Is the seniority right?
An AI session for a leadership team is not a shortened version of the team session. Executives need to make decisions about funding, sequencing, risk and accountability. Their teams need to produce work.
Mixing the two in one room usually serves neither. Ask whether the provider distinguishes them — and if the same deck appears for both, that is your answer.
8. What happens in the thirty days afterwards?
The failure mode is well documented: enthusiasm for two weeks, then reversion.
Ask what mechanism carries the learning forward. A written adoption plan per participant, a shared prompt library the team keeps, a named owner for each workflow, a review point in the calendar. “We’ll send the slides” is not a mechanism.
9. What is the evidence?
Named clients, participant numbers, ratings. Not “trained hundreds of professionals” — which client, how many people, what did they say.
Where a provider can point to repeat engagements, that is stronger evidence than any satisfaction score. A first round can be sold on enthusiasm. A second has to be earned.
The order matters more than the provider
One structural point, independent of who you choose.
Start with the leadership group, not with a departmental pilot. Departmental training without leadership involvement produces pockets of enthusiasm that fade, because nobody in the room controls priorities or budget.
When the leadership team has a shared position first, departmental training lands into an organisation that has already decided what it is trying to achieve.
The honest caveat: it can work in the other order. At Kahraba the foundation workshop ran three times — 65 executives, rated 4.8 out of 5 — and the CEO and senior managers booked their own advanced session afterwards, once the organisation had proved what it could do. What is not optional either way is the data boundary at the front.
Once you have settled what to buy, the next question is what it should cost and how it is quoted. That is answered separately in what corporate AI training costs in Egypt and the Gulf.
Ahmed Amrousy is an Egyptian AI-for-Business Expert, Executive Education Instructor at the American University in Cairo’s Onsi Sawiris School of Business, Head of Marketing & PR at Nile Air, and the founder of Menova.