# The questions regional CEOs keep asking me about AI

- Link: https://xanadu.co/questions-regional-ceos-ask-about-ai/
- Published: 2026-07-09T10:05:40+04:00
- Author: Daragh McGrath

Most weeks, I’m across the table from a CEO somewhere in the Gulf. Real estate, 
financial services, government-adjacent businesses. A year ago, the AI conversation
was polite and abstract: what is this, and does it matter? Now it’s sharper, and
a bit more uncomfortable: what do I actually do about it, and am I already behind?

The questions are remarkably consistent, whatever the sector or the size of the 
business. Here are the ones that come up in almost every conversation, and honest
answers. Not the ones a vendor gives you.

## “Is it too early, or am I already behind?”

Neither, and both. The honest answer is that the real risk was never moving too 
early or too late. It’s moving without a plan. Spending on tools because the board
asked, and having nothing to show for it a year later.

The region adds its own pressure here. When your government has a national AI strategy
and a minister to match, the top-down expectation is real, and it’s felt in every
boardroom. But “doing AI” and “buying AI” are not the same thing, and the gap between
them is exactly where most budgets quietly disappear. The window to build a genuine
advantage is open now. It closes not when the technology matures, but when your 
competitors work out how to operationalise it, and you still haven’t.

## “Where do I start?”

Not with a tool. That’s the most common and most expensive mistake there is: picking
a platform first, then hunting for a problem to justify it.

Start with one workflow where the value is obvious and the risk is low. One process
your team does every week that’s slow, manual and repeatable. Prove it works there,
in your business, with your people. Then take the pattern and repeat it somewhere
else. The companies that get stuck are the ones that tried to transform everything
at once, or ran a proof of concept so small it proved nothing. Start narrow, prove
it, scale the pattern. That’s the whole method.

## “How do I use this without losing control of my brand and my data?”

This is the question I get most from businesses that have something to protect, 
which, around here, is most of them. A regulated balance sheet. A brand built carefully
over decades in a market where aesthetics matter. Client data that cannot leave 
the building.

The mistake is treating it as a choice: move fast and risk it, or stay safe and 
stall. It isn’t a choice. You build the guardrails so the AI only ever works with
approved, on-brand, permitted material, and then you let it run inside them. Ungoverned
AI that generates whatever it likes isn’t a feature; it’s a liability with a friendly
interface. The businesses doing this well aren’t the most cautious ones. They’re
the ones who set the boundaries first, then move quickly inside them.

## “Will this replace my team?”

Almost never in the way people fear. The value in practice isn’t fewer people. It’s
your existing people spending less time on the work that drains them and more on
the work only they can do.

Here’s what’s already happening in most companies, whether leadership knows it or
not. People are using AI quietly, in browser tabs nobody can see, getting real value
nobody can measure. That isn’t adoption. It’s experimentation hiding in the org 
chart. The job isn’t to replace anyone. It’s to bring that into the open, make it
safe, and turn it into something the whole team can rely on. And in a region investing
this heavily in developing its own talent, AI that makes your people more capable
is worth far more than AI that makes them redundant.

## “Everyone is selling me AI. How do I know who’s real?”

Fair question, because the market is flooded. Every consultancy, every agency, every
software vendor has an AI story now, and most of them sound the same.

Two filters cut through it. First, ask for proof, not a demo. A polished demo shows
what the technology can do in ideal conditions. Proof shows what it did in a business
like yours. Second, ask who’s still in the room when it gets hard. Plenty of firms
will fly in, run a brilliant workshop, and fly out with the invoice. The actual 
work of landing AI in your building, with your team, against your regulator, in 
your language, happens over months and on the ground, with someone who’s still there
when the first thing breaks. That part doesn’t happen over a video call from another
timezone. And it’s the part that decides whether any of this sticks.

## The thread through all of it

If there’s one line under every one of these questions, it’s this. The CEOs who 
win the next few years won’t be the ones who bought the most AI. They’ll be the 
ones who got it out of the experiment stage and into how the business actually runs.
Deliberately, safely, and with the right people by their side.

That’s the whole game. Not the technology. The operationalising of it.

---

_These are the questions we spend our days answering with businesses across the 
region. If they’re the ones on your desk, [let’s talk](https://xanadu.co/contact/)._